LVLMs have been shown to perform excellently in image-level tasks such as VQA and caption. However, in many instance-level tasks, such as visual grounding and object detection, LVLMs still show performance gaps compared to previous expert models. Meanwhile, although pedestrian tracking is a classical task, there have been a number of new topics in combining object tracking and natural language, such as Referring MOT, Cross-view Referring MOT, and Semantic MOT. These tasks emphasize that models should understand the tracked object at an advanced semantic level, which is exactly where LVLMs excel. In this paper, we propose a new unified Pedestrian Tracking framework, namely OmniPT, which can track, track based on reference and generate semantic understanding of tracked objects interactively. We address two issues: how to model the tracking task into a task that foundation models can perform, and how to make the model output formatted answers. To this end, we implement a training phase consisting of RL-Mid Training-SFT-RL. Based on the pre-trained weights of the LVLM, we first perform a simple RL phase to enable the model to output fixed and supervisable bounding box format. Subsequently, we conduct a mid-training phase using a large number of pedestrian-related datasets. Finally, we perform supervised fine-tuning on several pedestrian tracking datasets, and then carry out another RL phase to improve the model's tracking performance and enhance its ability to follow instructions. We conduct experiments on tracking benchmarks and the experimental results demonstrate that the proposed method can perform better than the previous methods.

Computer-Aided Design (CAD) plays a vital role in engineering and manufacturing, yet current CAD workflows require extensive domain expertise and manual modeling effort. Recent advances in large language models (LLMs) have made it possible to generate code from natural language, opening new opportunities for automating parametric 3D modeling. However, directly translating human design intent into executable CAD code remains highly challenging, due to the need for logical reasoning, syntactic correctness, and numerical precision. In this work, we propose CAD-RL, a multimodal Chain-of-Thought (CoT) guided reinforcement learning post training framework for CAD modeling code generation. Our method combines CoT-based Cold Start with goal-driven reinforcement learning post training using three task-specific rewards: executability reward, geometric accuracy reward, and external evaluation reward. To ensure stable policy learning under sparse and high-variance reward conditions, we introduce three targeted optimization strategies: Trust Region Stretch for improved exploration, Precision Token Loss for enhanced dimensions parameter accuracy, and Overlong Filtering to reduce noisy supervision. To support training and benchmarking, we release ExeCAD, a noval dataset comprising 16,540 real-world CAD examples with paired natural language and structured design language descriptions, executable CADQuery scripts, and rendered 3D models. Experiments demonstrate that CAD-RL achieves significant improvements in reasoning quality, output precision, and code executability over existing VLMs.

Existing stereotype auditing methods for Large Language Models (LLMs) typically rely on isolated rating schemes or task-specific probes, lacking theoretical grounding and failing to reveal internal organization beyond surface-level output patterns. In this paper, we introduce SCoUT (Stereotype Content-oriented Utility structure via Thurstonian modeling), a closed-loop framework that structurally models, explicitly probes, and functionally steers stereotype dimensions (warmth and competence) in LLMs. SCoUT first reconstructs a global stereotype utility structure aligned with Stereotype Content Model theory via Thurstonian comparative judgments. Across multiple open-source LLMs, this modeling achieves high pairwise-preference prediction accuracy (≥ 0.90 on larger-scale models) and exhibits strong cross-model consistency. Probing internal attention mechanisms localizes this structure to specific heads (Spearman’s ρ up to 0.83 for warmth and 0.90 for competence) and surfaces a salient asymmetry between warmth and competence. Further, targeted inference-time activation modifications on these dimension-sensitive heads consistently steer model outputs along the intended axes. By bridging behavioral measurement with internal representation and controllable steering, SCoUT offers an end-to-end framework that uncovers and interprets the latent structure of stereotypes, advancing stereotype auditing from surface detection to structural analysis.

Stochastic partition processes divide a multi-dimensional space into a number of regions, such that the data within each region exhibit some form of homogeneity. Due to the nature of their partition strategies, partition processes can often create many unnecessary divisions in sparse regions when trying to describe data in dense regions. To avoid this problem we introduce a parsimonious partition model - the Rectangular Bounding Process (RBP) - to efficiently partition multi-dimensional spaces, by employing a bounding strategy to enclose data points within rectangular bounding boxes. The RBP is self-consistent and as such can be directly extended from a finite hypercube to an infinite (unbounded) space. We extend the RBP to establish a data-dependent RBP (data-RBP) to generate bounding boxes only over existing data points in a sequential manner, which can effectively reduce model complexity and enable online learning. To achieve this, we design an alternative way to generate bounding boxes and prove the distributional equivalence between the data-RBP and the RBP when empty boxes are removed. We demonstrate application of the RBP and the data-RBP in three scenarios: regression trees, relational modelling, and random feature construction for online learning. Extensive experimental results validate the performance of the RBP and the data-RBP for both accuracy and efficiency.

Video anomaly detection (VAD) has been widely studied for its important applications in multimedia community. Recently, many Weakly Supervised VAD (WS-VAD) methods have been proposed, which tend to treat VAD as a classification task through multiple instance learning and result in the need to collect sufficient anomaly classes and samples to be used for training a classifier. However, anomaly events tend to be open-set and rare in real-world applications, so we often have difficulty collecting all anomaly classes and enough sample anomalies, which is a difficult situation for WS-VAD to cope with. To this end, we consider to treat VAD as an out-of-distribution detection task rather than a classification task and propose a simple but effective semi-supervised baseline method. First, we leverage the powerful zero-shot capability of large visual language models to generate summary text descriptions for videos and extract visual features as intermediates for subsequent use. Next, we use a text encoder to extract language features and combine them with visual features to obtain robust multimodal features. Finally, we introduce an out-of-distribution detection method that learns the center of normality in multimodal space from normal and unlabeled samples, while deviating abnormal samples from the center to cope with the scarcity of abnormal samples. To implement our baseline method, we also provide a new semi-supervised dataset by reorganizing an existing benchmark, which is the first available dataset in the VAD community that provides trimmed videos consisting of complete abnormal events. Experiments demonstrate that our method performs more robustly when fewer anomaly classes and anomaly samples collected.

Multiple Object Tracking (MOT) is a fundamental task in computer vision. Existing methods utilize motion information or appearance information to perform object tracking. However, these algorithms still struggle with special circumstances, such as occlusion and blurring in complex scenes. Inspired by the fact that people can pinpoint objects through verbal descriptions, we explore performing long-term robust tracking using semantic features of objects. Motivated by the success of the multimodal foundation model in text-image alignment, we reconsider the appearance feature extraction module in MOT and propose a Foundation model Driven multi-object tracker (FDTracker). Specifically, we propose a two-stage trained appearance feature extractor. In the first stage, using a single image of the object as input, the model could capture the attributes of objects with the assistance of natural language instructions. In the second stage, using a sequence of images of objects as input, the model learns how to use these attributes to distinguish between different objects and connect the same object at different times. Finally, for coordinating appearance and motion information, we propose a reasonable combined strategy, which better facilitates trajectory assignment and reconnection. Extensive experiments on benchmarks demonstrate the robustness of FDTracker.

Cloth-changing person re-identification aims at recognizing the same person with clothing changes across non-overlapping cameras. Advanced methods either resort to identity-related auxiliary modalities (e.g., sketches, silhouettes, and keypoints) or clothing labels to mitigate the impact of clothes. However, relying on unpractical and inflexible auxiliary modalities or annotations limits their real-world applicability. In this paper, we promote cloth-changing person re-identification by leveraging abundant semantics present within pedestrian images, without the need for any auxiliaries. Specifically, we first propose a unified Semantics Mining and Refinement (SMR) module to extract robust identity-related content and salient semantics, mitigating interference from clothing appearances effectively. We further propose the Content and Salient Semantics Collaboration (CSSC) framework to collaborate and leverage various semantics, facilitating cross-parallel semantic interaction and refinement. Our proposed method achieves state-of-the-art performance on three cloth-changing benchmarks, demonstrating its superiority over advanced competitors. The code is available at https://github.com/QizaoWang/CSSC-CCReID.

Person re-identification (Re-ID) is a crucial task in computer vision, aiming to recognize individuals across non-overlapping camera views. While recent advanced vision-language models (VLMs) excel in logical reasoning and multi-task generalization, their applications in Re-ID tasks remain limited. They either struggle to perform accurate matching based on identity-relevant features or assist image-dominated branches as auxiliary semantics. In this paper, we propose a novel framework ChatReID, that shifts the focus towards a text-side-dominated retrieval paradigm, enabling flexible and interactive re-identification. To integrate the reasoning abilities of language models into Re-ID pipelines, We first present a large-scale instruction dataset, which contains more than 8 million prompts to promote the model fine-tuning. Next. we introduce a hierarchical progressive tuning strategy, which endows Re-ID ability through three stages of tuning, i.e., from person attribute understanding to fine-grained image retrieval and to multi-modal task reasoning. Extensive experiments across ten popular benchmarks demonstrate that ChatReID outperforms existing methods, achieving state-of-the-art performance in all Re-ID tasks. More experiments demonstrate that ChatReID not only has the ability to recognize fine-grained details but also to integrate them into a coherent reasoning process.

Abstract visual reasoning (AVR) enables humans to quickly discover and generalize abstract rules to new scenarios. Designing intelligent systems with human-like AVR abilities has been a long-standing topic in the artificial intelligence community. Deep AVR solvers have recently achieved remarkable success in various AVR tasks. However, they usually use task-specific designs or parameters in different tasks. In such a paradigm, solving new tasks often means retraining the model, and sometimes retuning the model architectures, which increases the cost of solving AVR problems. In contrast to task-specific approaches, this paper proposes a novel Unified Conditional Generative Solver (UCGS), aiming to address multiple AVR tasks in a unified framework. First, we prove that some well-known AVR tasks can be reformulated as the problem of estimating the predictability of target images in problem panels. Then, we illustrate that, under the proposed framework, training one conditional generative model can solve various AVR tasks. The experiments show that with a single round of multi-task training, UCGS demonstrates abstract reasoning ability across various AVR tasks. Especially, UCGS exhibits the ability of zero-shot reasoning, enabling it to perform abstract reasoning on problems from unseen AVR tasks in the testing phase.

In recent years, One-shot Federated Learning (OSFL) methods based on Diffusion Models (DMs) have garnered increasing attention due to their remarkable performance. However, most of these methods require the deployment of foundation models on client devices, which significantly raises the computational requirements and reduces their adaptability to heterogeneous client models. In this paper, we propose FedLMG, a heterogeneous one-shot Federated learning method with Local Model-Guided diffusion models. In our method, clients do not need access to any foundation models but only train and upload their local models, which is consistent with traditional FL methods. On the clients, we employ classification loss and batch normalization loss to capture the broad category features and detailed contextual features of the client distributions. On the server, based on the uploaded client models, we utilize backpropagation to guide the server’s DM in generating synthetic datasets that comply with the client distributions, which are then used to train the aggregated model. By using the local models as a medium to transfer client knowledge, our method significantly reduces the computational requirements on client devices and effectively adapts to scenarios with heterogeneous clients. Extensive quantitation and visualization experiments on three large-scale real-world datasets, along with theoretical analysis, demonstrate that the synthetic datasets generated by FedLMG exhibit comparable quality and diversity to the client datasets, which leads to an aggregated model that outperforms all compared methods and even the performance ceiling, further elucidating the significant potential of utilizing DMs in FL.

Computer-Aided Design (CAD) is pivotal in industrial manufacturing, with orthographic projection reasoning foundational to its entire workflow—encompassing design, manufacturing, and simulation. However, prevailing deep-learning approaches employ standard 3D reconstruction pipelines as an alternative, which often introduce imprecise dimensions and limit the parametric editability required for CAD workflows. Recently, some researchers adopt vision–language models (VLMs), particularly supervised fine-tuning (SFT), to tackle CAD-related challenges. SFT shows promise but often devolves into pattern memorization, resulting in poor out-of-distribution (OOD) performance on complex reasoning tasks. To tackle these limitations, we introduce CReFT-CAD, a two-stage fine-tuning paradigm: first, a curriculum-driven reinforcement learning stage with difficulty-aware rewards to steadily build reasoning abilities; second, supervised post-tuning to refine instruction following and semantic extraction. Complementing this, we release TriView2CAD, the first large-scale, open-source benchmark for orthographic projection reasoning, comprising 200,000 synthetic and 3,000 real-world orthographic projections with precise dimensional annotations and six interoperable data modalities. Benchmarking leading VLMs on orthographic projection reasoning, we show that CReFT-CAD significantly improves reasoning accuracy and OOD generalizability in real-world scenarios, providing valuable insights to advance CAD reasoning research. The code and adopted datasets are available at \url{https://github.com/KeNiu042/CReFT-CAD}.

With the continuous expansion of intelligent surveillance networks, lifelong person re-identification (LReID) has received widespread attention, pursuing the need of self-evolution across different domains. However, existing LReID studies accumulate knowledge with the assumption that people would not change their clothes. In this paper, we propose a more practical task, namely lifelong person re-identification with hybrid clothing states (LReID-Hybrid), which takes a series of cloth-changing and same-cloth domains into account during lifelong learning. To tackle the challenges of knowledge granularity mismatch and knowledge presentation mismatch in LReID-Hybrid, we take advantage of the consistency and generalization capabilities of the text space, and propose a novel framework, dubbed Teata, to effectively align, transfer, and accumulate knowledge in an "image-text-image" closed loop. Concretely, to achieve effective knowledge transfer, we design a Structured Semantic Prompt (SSP) learning to decompose the text prompt into several structured pairs to distill knowledge from the image space with a unified granularity of text description. Then, we introduce a Knowledge Adaptation and Projection (KAP) strategy, which tunes text knowledge via a slow-paced learner to adapt to different tasks without catastrophic forgetting. Extensive experiments demonstrate the superiority of our proposed Teata for LReID-Hybrid as well as on conventional LReID benchmarks over advanced methods.

The ability to extract invariant visual features of objects from complex scenes and identify the same objects in different scenes is inborn for humans. To endow AI systems with such capability, we introduce a novel compositional scene understanding method known as Compositional Scene understanding via Global Object-centric representations (CSGOs). CSGO achieves comprehensive scene understanding, including the discovery and identification of objects, by leveraging a set of learnable global object-centric representations in an unsupervised manner. CSGO comprises three components: 1) Local Object-Centric Learning, which is responsible for extracting localized and scene-specific object-centric representations to discover objects; 2) Image Decoding, facilitating the reconstruction of object and scene images using the obtained object-centric representation as input; and 3) Global Object-Centric Learning, identifying the object across diverse scenes according to a set of learnable global object-centric representations, which indicates the scene-free intrinsic attributes (i.e., appearance and shape) of objects. Experimental results on three synthetic datasets and one real-world scene dataset demonstrate that CSGO has excellent object identification and attribute disentanglement abilities. Furthermore, the scene decomposition performance (indicating object discovery performance) of CSGO is superior to comparison methods.

Federated learning (FL) enables multiple clients to collaboratively train a global model without disclosing their data. Previous researches often require training the complete model parameters. However, the emergence of powerful pre-trained models makes it possible to achieve higher performance with fewer learnable parameters in FL. In this paper, we propose a federated adaptive prompt tuning algorithm, FedAPT, for multi-domain collaborative image classification with powerful foundation models, like CLIP. Compared with direct federated prompt tuning, our core idea is to adaptively unlock specific domain knowledge for each test sample in order to provide them with personalized prompts. To implement this idea, we design an adaptive prompt tuning module, which consists of a meta prompt, an adaptive network, and some keys. The server randomly generates a set of keys and assigns a unique key to each client. Then all clients cooperatively train the global adaptive network and meta prompt with the local datasets and the frozen keys. Ultimately, the global aggregation model can assign a personalized prompt to CLIP based on the domain features of each test sample. We perform extensive experiments on two multi-domain image classification datasets across two different settings -- supervised and unsupervised. The results show that FedAPT can achieve better performance with less than 10% of the number of parameters of the fully trained model, and the global model can perform well in diverse client domains simultaneously.

Recently, semi-supervised federated learning (semi-FL) has been proposed to handle the commonly seen real-world scenarios with labeled data on the server and unlabeled data on the clients. However, existing methods face several challenges such as communication costs, data heterogeneity, and training pressure on client devices. To address these challenges, we introduce the powerful diffusion models (DM) into semi-FL and propose FedDISC, a Federated Diffusion-Inspired Semi-supervised Co-training method. Specifically, we first extract prototypes of the labeled server data and use these prototypes to predict pseudo-labels of the client data. For each category, we compute the cluster centroids and domain-specific representations to signify the semantic and stylistic information of their distributions. After adding noise, these representations are sent back to the server, which uses the pre-trained DM to generate synthetic datasets complying with the client distributions and train a global model on it. With the assistance of vast knowledge within DM, the synthetic datasets have comparable quality and diversity to the client images, subsequently enabling the training of global models that achieve performance equivalent to or even surpassing the ceiling of supervised centralized training. FedDISC works within one communication round, does not require any local training, and involves very minimal information uploading, greatly enhancing its practicality. Extensive experiments on three large-scale datasets demonstrate that FedDISC effectively addresses the semi-FL problem on non-IID clients and outperforms the compared SOTA methods. Sufficient visualization experiments also illustrate that the synthetic dataset generated by FedDISC exhibits comparable diversity and quality to the original client dataset, with a neglectable possibility of leaking privacy-sensitive information of the clients.

Endowing machines with abstract reasoning ability has been a long-term research topic in artificial intelligence. Raven's Progressive Matrix (RPM) is widely used to probe abstract visual reasoning in machine intelligence, where models will analyze the underlying rules and select one image from candidates to complete the image matrix. Participators of RPM tests can show powerful reasoning ability by inferring and combining attribute-changing rules and imagining the missing images at arbitrary positions of a matrix. However, existing solvers can hardly manifest such an ability in realistic RPM tests. In this paper, we propose a deep latent variable model for answer generation problems through Rule AbstractIon and SElection (RAISE). RAISE can encode image attributes into latent concepts and abstract atomic rules that act on the latent concepts. When generating answers, RAISE selects one atomic rule out of the global knowledge set for each latent concept to constitute the underlying rule of an RPM. In the experiments of bottom-right and arbitrary-position answer generation, RAISE outperforms the compared solvers in most configurations of realistic RPM datasets. In the odd-one-out task and two held-out configurations, RAISE can leverage acquired latent concepts and atomic rules to find the rule-breaking image in a matrix and handle problems with unseen combinations of rules and attributes.

In recent years, the attention towards One-Shot Federated Learning (OSFL) has been driven by its capacity to minimize communication. With the development of the diffusion model (DM), several methods employ the DM for OSFL, utilizing model parameters, image features, or textual prompts as mediums to transfer the local client knowledge to the server. However, these mediums often require public datasets or the uniform feature extractor, significantly limiting their practicality. In this paper, we propose FedDEO, a Description-Enhanced One-Shot Federated Learning Method with DMs, offering a novel exploration of utilizing the DM in OSFL. The core idea of our method involves training local descriptions on the clients, serving as the medium to transfer the knowledge of the distributed clients to the server. Firstly, we train local descriptions on the client data to capture the characteristics of client distributions, which are then uploaded to the server. On the server, the descriptions are used as conditions to guide the DM in generating synthetic datasets that comply with the distributions of various clients, enabling the training of the aggregated model. Theoretical analyses and sufficient quantitation and visualization experiments on three large-scale real-world datasets demonstrate that through the training of local descriptions, the server is capable of generating synthetic datasets with high quality and diversity. Consequently, with advantages in communication and privacy protection, the aggregated model outperforms compared FL or diffusion-based OSFL methods and, on some clients, outperforms the performance ceiling of centralized training.

Given the complexities inherent in visual scenes, such as object occlusion, a comprehensive understanding often requires observation from multiple viewpoints. Existing multi-viewpoint object-centric learning methods typically employ random or sequential viewpoint selection strategies. While applicable across various scenes, these strategies may not always be ideal, as certain scenes could benefit more from specific viewpoints. To address this limitation, we propose a novel active viewpoint selection strategy. This strategy predicts images from unknown viewpoints based on information from observation images for each scene. It then compares the object-centric representations extracted from both viewpoints and selects the unknown viewpoint with the largest disparity, indicating the greatest gain in information, as the next observation viewpoint. Through experiments on various datasets, we demonstrate the effectiveness of our active viewpoint selection strategy, significantly enhancing segmentation and reconstruction performance compared to random viewpoint selection. Moreover, our method can accurately predict images from unknown viewpoints.

Visual scenes are extremely diverse, not only because there are infinite possible combinations of objects and backgrounds but also because the observations of the same scene may vary greatly with the change of viewpoints. When observing a multi-object visual scene from multiple viewpoints, humans can perceive the scene compositionally from each viewpoint while achieving the so-called "object constancy" across different viewpoints, even though the exact viewpoints are untold. This ability is essential for humans to identify the same object while moving and to learn from vision efficiently. It is intriguing to design models that have a similar ability. In this article, we consider a novel problem of learning compositional scene representations from multiple unspecified (i.e., unknown and unrelated) viewpoints without using any supervision and propose a deep generative model which separates latent representations into a viewpoint-independent part and a viewpoint-dependent part to solve this problem. During the inference, latent representations are randomly initialized and iteratively updated by integrating the information in different viewpoints with neural networks. Experiments on several specifically designed synthetic datasets have shown that the proposed method can effectively learn from multiple unspecified viewpoints.

Given a network, it is well recognized that attributed network embedding represents each node of the network in a low-dimensional space, and, thus, brings considerable benefits for numerous graph mining tasks. In practice, a diverse set of graph tasks can be processed efficiently via the compact representation that preserves content and structure information. The majority of attributed network embedding approaches, especially, the graph neural network (GNN) algorithms, are substantially costly in either time or space due to the expensive learning process, while the randomized hashing technique, locality-sensitive hashing (LSH), which does not need learning, can speedup the embedding process at the expense of losing some accuracy. In this article, we propose the MPSketch model, which bridges the performance gap between the GNN framework and the LSH framework by adopting the LSH technique to pass messages and capture high-order proximity in a larger aggregated information pool from the neighborhood. The extensive experimental results confirm that in node classification and link prediction, the proposed MPSketch algorithm enjoys performance comparable to the state-of-the-art learning-based algorithms and outperforms the existing LSH algorithms, while running faster than the GNN algorithms by 3-4 orders of magnitude. More precisely, MPSketch runs 2121, 1167, and 1155 times faster than GraphSAGE, GraphZoom, and FATNet on average, respectively.

Cloth-changing person Re-IDentification (Re-ID) is a particularly challenging task, suffering from two limitations of inferior discriminative features and limited training samples. Existing methods mainly leverage auxiliary information to facilitate identity-relevant feature learning, including soft-biometrics features of shapes or gaits, and additional labels of clothing. However, this information may be unavailable in real-world applications. In this paper, we propose a novel FIne-grained Representation and Recomposition (FIRe2) framework to tackle both limitations without any auxiliary annotation or data. Specifically, we first design a Fine-grained Feature Mining (FFM) module to separately cluster images of each person. Images with similar so-called fine-grained attributes (e.g., clothes and viewpoints) are encouraged to cluster together. An attribute-aware classification loss is introduced to perform fine-grained learning based on cluster labels, which are not shared among different people, promoting the model to learn identity-relevant features. Furthermore, to take full advantage of fine-grained attributes, we present a Fine-grained Attribute Recomposition (FAR) module by recomposing image features with different attributes in the latent space. It significantly enhances robust feature learning. Extensive experiments demonstrate that FIRe2 can achieve state-of-the-art performance on five widely-used cloth-changing person Re-ID benchmarks. The code is available athttps://github.com/QizaoWang/FIRe-CCReID.

Scene text recognition has been studied for decades due to its broad applications. However, despite Chinese characters possessing different characteristics from Latin characters, such as complex inner structures and large categories, few methods have been proposed for Chinese Text Recognition (CTR). Particularly, the characteristic of large categories poses challenges in dealing with zero-shot and few-shot Chinese characters. In this paper, inspired by the way humans recognize Chinese texts, we propose a two-stage framework for CTR. Firstly, we pre-train a CLIP-like model through aligning printed character images and Ideographic Description Sequences (IDS). This pre-training stage simulates humans recognizing Chinese characters and obtains the canonical representation of each character. Subsequently, the learned representations are employed to supervise the CTR model, such that traditional single-character recognition can be improved to text-line recognition through image-IDS matching. To evaluate the effectiveness of the proposed method, we conduct extensive experiments on both Chinese character recognition (CCR) and CTR. The experimental results demonstrate that the proposed method performs best in CCR and outperforms previous methods in most scenarios of the CTR benchmark. It is worth noting that the proposed method can recognize zero-shot Chinese characters in text images without fine-tuning, whereas previous methods require fine-tuning when new classes appear. The code is available at https://github.com/FudanVI/FudanOCR/tree/main/image-ids-CTR.

Data similarity computation is a fundamental research topic which underpins many high-level applications based on similarity measures. However, the exact similarity computation has become daunting in large-scale real-world scenarios. Currently, MinHash is a popular technique for efficiently estimating the Jaccard similarity of binary sets and, furthermore, weighted MinHash is utilized to estimate the generalized Jaccard similarity of weighted sets. This review focuses on categorizing and discussing the existing works of weighted MinHash algorithms. Also, we have developed a Python toolbox for the algorithms, and released it in our github.

Detection models trained by one party (including server) may face severe performance degradation when distributed to other users (clients). Federated learning can enable multi-party collaborative learning without leaking client data. In this paper, we focus on a special cross-domain scenario in which the server has large-scale labeled data and multiple clients only have a small amount of labeled data; meanwhile, there exist differences in data distributions among the clients. In this case, traditional federated learning methods can’t help a client learn both the global knowledge of all participants and its own unique knowledge. To make up for this limitation, we propose a cross-domain federated object detection framework, named FedOD. The proposed framework first performs the federated training to obtain a public global aggregated model through multi-teacher distillation, and sends the aggregated model back to each client for fine-tuning its personalized local model. After a few rounds of communication, on each client we can perform weighted ensemble inference on the public global model and the personalized local model. We establish a federated object detection dataset which has significant background differences and instance differences based on multiple public autonomous driving datasets, and then conduct extensive experiments on the dataset. The experimental results validate the effectiveness of the proposed method.

In recent years, deep learning techniques have made significant advancements in text segmentation. However, most existing methods do not take into account that characters are composed of smaller components, such as strokes and other local patterns. Furthermore, the similarities between text components are crucial for effective text segmentation. With this in mind, we propose a multi-level Transformer-based method for text segmentation that incorporates a recognition module. To enhance the interaction between text components and extract features at different granularities, we introduce Global and Local Self-Attention blocks. Our recognition module is trained jointly with the segmentation module to improve the model’s ability to focus on text details and improve its perception of texts. By aggregating features from multiple granularities, our segmentation module produces accurate pixel-level mask predictions. The experimental results demonstrate the effectiveness of our approach on several text segmentation benchmarks and show that it outperforms existing methods.

Scene text recognition (STR) has attracted much attention due to its broad applications. The previous works pay more attention to dealing with the recognition of Latin text images with complex backgrounds by introducing language models or other auxiliary networks. Different from Latin texts, many vertical Chinese texts exist in natural scenes, which brings difficulties to current state-of-the-art STR methods. In this paper, we take the first attempt to extract orientation-independent visual features by disentangling content and orientation information of text images, thus recognizing both horizontal and vertical texts robustly in natural scenes. Specifically, we introduce a Character Image Reconstruction Network (CIRN) to recover corresponding printed character images with disentangled content and orientation information. We conduct experiments on a scene dataset for benchmarking Chinese text recognition, and the results demonstrate that the proposed method can indeed improve performance through disentangling content and orientation information. To further validate the effectiveness of our method, we additionally collect a Vertical Chinese Text Recognition (VCTR) dataset. The experimental results show that the proposed method achieves 45.63\% improvement on VCTR when introducing CIRN to the baseline model.

Video text spotting (VTS) aims at extracting texts from videos, where text detection, tracking and recognition are conducted simultaneously. There have been some works that can tackle VTS; however, they may ignore the underlying semantic relationships among texts within a frame. We observe that the texts within a frame usually share similar semantics, which suggests that, if one text is predicted incorrectly by a text recognizer, it still has a chance to be corrected via semantic reasoning. In this paper, we propose an accurate video text spotter, VLSpotter, that reads texts visually, linguistically, and semantically. For ‘visually’, we propose a plug-and-play text-focused super-resolution module to alleviate motion blur and enhance video quality. For ‘linguistically’, a language model is employed to capture intra-text context to mitigate wrongly spelled text predictions. For ‘semantically’, we propose a text-wise semantic reasoning module to model inter-text semantic relationships and reason for better results. The experimental results on multiple VTS benchmarks demonstrate that the proposed VLSpotter outperforms the existing state-of-the-art methods in end-to-end video text spotting.

Multiple object tracking (MOT) tends to become more challenging when severe occlusions occur. In this paper, we analyze the limitations of traditional Convolutional Neural Network-based methods and Transformer-based methods in handling occlusions and propose DNMOT, an end-to-end trainable DeNoising Transformer for MOT. To address the challenge of occlusions, we explicitly simulate the scenarios when occlusions occur. Specifically, we augment the trajectory with noises during training and make our model learn the denoising process in an encoder-decoder architecture, so that our model can exhibit strong robustness and perform well under crowded scenes. Additionally, we propose a Cascaded Mask strategy to better coordinate the interaction between different types of queries in the decoder to prevent the mutual suppression between neighboring trajectories under crowded scenes. Notably, the proposed method requires no additional modules like matching strategy and motion state estimation in inference. We conduct extensive experiments on the MOT17, MOT20, and DanceTrack datasets, and the experimental results show that our method outperforms previous state-of-the-art methods by a clear margin.

Text segmentation is a crucial aspect of various text-related tasks, including text erasing, text editing, and font style transfer. In recent years, multiple text segmentation datasets, such as TextSeg focusing on Latin text segmentation and BTS on bilingual text segmentation, have been proposed. However, existing methods either disregard the annotations of text location or directly use pre-trained text detectors. In general, these methods cannot fully utilize the annotations of text location in the datasets. To explicitly incorporate text location information to guide text segmentation, we propose an end-to-end text-focused segmentation framework, where text detection and segmentation are jointly optimized. In the proposed framework, we first extract multi-level global visual features through residual convolution blocks and then predict the mask of text areas using a text detection head. Subsequently, we develop a text-focused module that compels the model to pay more attention to text areas. Specifically, we introduce two types of attention masks to extract corresponding features: text-aware and instance-aware features. Finally, we employ hierarchical Transformer encoders to fuse multi-level features and predict the text mask with a text segmentation head. To evaluate the effectiveness of our method, we conduct experiments on six text segmentation benchmarks. The experimental results demonstrate that the proposed method outperforms the previous state-of-the-art (SOTA) methods by a clear margin in most cases. The code and supplementary materials are available at https://github.com/FudanVI/FudanOCR/tree/main/text-focused-Transformers https://github.com/FudanVI/FudanOCR/tree/main/text-focused-Transformers.

Text segmentation is a challenging computer vision task with many downstream applications. Current text segmentation models need to be trained with pixel-level annotations, which requires a lot of labor cost. In this paper, we take the first attempt to perform weakly-supervised text instance segmentation through bridging text recognition and text segmentation. We observe that text recognition models are able to produce the attention localization of each text instance. Based on this observation, we propose a two-stage Text Adaptive Refinement (TAR) module to generate the pseudo labels based on the attention map of a text recognizer. Meanwhile, we develop a text segmentation module to take the rough attention location as input to predict segmentation masks, which are supervised by the aforementioned pseudo labels. In addition, we introduce a mask-augmented contrastive learning by treating the segmentation result as an augmented version of the input text image, thus improving the visual representation and further enhancing the performance of both recognition and segmentation. The experimental results demonstrate that the proposed method outperforms the state-of-the-art (SOTA) weakly-supervised generic segmentation methods by 18.95% and 17.80% in fgIoU on ICDAR13-FST and TextSeg. On MLT-S, COCO-TS and Total-Text, the proposed method achieves about 82% of the fully-supervised methods' performance. When evaluated on instance segmentation, the proposed method exceeds existing SOTA methods by 23.32% and 21.34% on ICDAR13-FST and TextSeg, respectively. Code and Supplementary Materials are available at https://github.com/FudanVI/FudanOCR/tree/main/weakly-text-segmentation.

Diffusion models (DMs) have recently gained attention with state-of-the-art performance in text-to-image synthesis. Abiding by the tradition in deep learning, DMs are trained and evaluated on the images with fixed sizes. However, users are demanding for various images with specific sizes and various aspect ratio. This paper focuses on adapting text-to-image diffusion models to handle such variety while maintaining visual fidelity. First we observe that, during the synthesis, lower resolution images suffer from incomplete object portrayal, while higher resolution images exhibit repetitively disordered presentation. Next, we establish a statistical relationship indicating that attention entropy changes with token quantity, suggesting that models aggregate spatial information in proportion to image resolution. The subsequent interpretation on our observations is that objects are incompletely depicted due to limited spatial information for low resolutions, while repetitively disorganized presentation arises from redundant spatial information for high resolutions. From this perspective, we propose a scaling factor to alleviate the change of attention entropy and mitigate the defective pattern observed. Extensive experimental results validate the efficacy of the proposed scaling factor, enabling models to achieve better visual effects, image quality, and text alignment. Notably, these improvements are achieved without additional training or fine-tuning techniques.

When perceiving the world from multiple viewpoints, humans have the ability to reason about the complete objects in a compositional manner even when an object is completely occluded from certain viewpoints. Meanwhile, humans are able to imagine novel views after observing multiple viewpoints. Recent remarkable advances in multi-view object-centric learning still leaves some unresolved problems: 1) The shapes of partially or completely occluded objects can not be well reconstructed. 2) The novel viewpoint prediction depends on expensive viewpoint annotations rather than implicit rules in view representations. In this paper, we introduce a time-conditioned generative model for videos. To reconstruct the complete shape of an object accurately, we enhance the disentanglement between the latent representations of objects and views, where the latent representations of time-conditioned views are jointly inferred with a Transformer and then are input to a sequential extension of Slot Attention to learn object-centric representations. In addition, Gaussian processes are employed as priors of view latent variables for video generation and novel-view prediction without viewpoint annotations. Experiments on multiple datasets demonstrate that the proposed model can make object-centric video decomposition, reconstruct the complete shapes of occluded objects, and make novel-view predictions.

Continuous-time interaction data is usually generated under time-evolving environment. Hawkes processes (HP) are commonly used mechanisms for the analysis of such data. However, typical model implementations (such as e.g., stochastic block models) assume that the exogenous (background) interaction rate is constant, and so they are limited in their ability to adequately describe any complex time-evolution in the background rate of a process. In this paper, we introduce a stochastic exogenous rate Hawkes process (SE-HP) which is able to learn time variations in the exogenous rate. The model affiliates each node with a piecewise-constant membership distribution with an unknown number of changepoint locations, and allows these distributions to be related to the membership distributions of interacting nodes. The time-varying background rate function is derived through combinations of these membership functions. We introduce a stochastic gradient MCMC algorithm for efficient, scalable inference. The performance of the SE-HP is explored on real world, continuous-time interaction datasets, where we demonstrate that the SE-HP strongly outperforms comparable state-of-the-art methods.

Visual scenes are composed of visual concepts and have the property of combinatorial explosion. An important reason for humans to efficiently learn from diverse visual scenes is the ability of compositional perception, and it is desirable for artificial intelligence to have similar abilities. Compositional scene representation learning is a task that enables such abilities. In recent years, various methods have been proposed to apply deep neural networks, which have been proven to be advantageous in representation learning, to learn compositional scene representations via reconstruction, advancing this research direction into the deep learning era. Learning via reconstruction is advantageous because it may utilize massive unlabeled data and avoid costly and laborious data annotation. In this survey, we first outline the current progress on reconstruction-based compositional scene representation learning with deep neural networks, including development history and categorizations of existing methods from the perspectives of the modeling of visual scenes and the inference of scene representations; then provide benchmarks, including an open source toolbox to reproduce the benchmark experiments, of representative methods that consider the most extensively studied problem setting and form the foundation for other methods; and finally discuss the limitations of existing methods and future directions of this research topic.

In the last decade, the blossom of deep learning has witnessed the rapid development of scene text recognition. However, the recognition of low-resolution scene text images remains a challenge. Even though some super-resolution methods have been proposed to tackle this problem, they usually treat text images as general images while ignoring the fact that the visual quality of strokes (the atomic unit of text) plays an essential role for text recognition. According to Gestalt Psychology, humans are capable of composing parts of details into the most similar objects guided by prior knowledge. Likewise, when humans observe a low-resolution text image, they will inherently use partial stroke-level details to recover the appearance of holistic characters. Inspired by Gestalt Psychology, we put forward a Stroke-Aware Scene Text Image Super-Resolution method containing a Stroke-Focused Module (SFM) to concentrate on stroke-level internal structures of characters in text images. Specifically, we attempt to design rules for decomposing English characters and digits at stroke-level, then pre-train a text recognizer to provide stroke-level attention maps as positional clues with the purpose of controlling the consistency between the generated super-resolution image and high-resolution ground truth. The extensive experimental results validate that the proposed method can indeed generate more distinguishable images on TextZoom and manually constructed Chinese character dataset Degraded-IC13. Furthermore, since the proposed SFM is only used to provide stroke-level guidance when training, it will not bring any time overhead during the test phase. Code is available at https://github.com/FudanVI/FudanOCR/tree/main/text-gestalt.

Visual scenes are extremely rich in diversity, not only because there are infinite combinations of objects and background, but also because the observations of the same scene may vary greatly with the change of viewpoints. When observing a visual scene that contains multiple objects from multiple viewpoints, humans are able to perceive the scene in a compositional way from each viewpoint, while achieving the so-called ``object constancy'' across different viewpoints, even though the exact viewpoints are untold. This ability is essential for humans to identify the same object while moving and to learn from vision efficiently. It is intriguing to design models that have the similar ability. In this paper, we consider a novel problem of learning compositional scene representations from multiple unspecified viewpoints without using any supervision, and propose a deep generative model which separates latent representations into a viewpoint-independent part and a viewpoint-dependent part to solve this problem. To infer latent representations, the information contained in different viewpoints is iteratively integrated by neural networks. Experiments on several specifically designed synthetic datasets have shown that the proposed method is able to effectively learn from multiple unspecified viewpoints.

Chinese character recognition (CCR) has drawn continuous research interest due to its wide applications. After decades of study, there still exist several challenges,e.g., different characters with similar appearance and the one-to-many problem. There is no unified solution to the above challenges as previous methods tend to address these problems separately. In this paper, we propose a Chinese character recognition method named Augmented Character Profile Matching (ACPM), which utilizes a collection of character knowledge from three decomposition levels to recognize Chinese characters. Specifically, the feature maps of each character image are utilized as the character-level knowledge. In addition, we introduce a radical-stroke counting module (RSC) to help produce augmented character profiles, including the number of radicals, the number of strokes, and the total length of strokes, which characterize the character more comprehensively. The feature maps of the character image and the outputs of the RSC module are collected to constitute a character profile for selecting the closest candidate character through joint matching. The experimental results show that the proposed method outperforms the state-of-the-art methods on both the ICDAR 2013 and CTW datasets by 0.35% and 2.23%, respectively. Moreover, it also clearly outperforms the compared methods in the zero-shot settings. Code is available at https://github.com/FudanVI/FudanOCR/tree/main/character-profile-matching.

Data similarity (or distance) computation is a fundamental research topic which underpins many high-level applications based on similarity measures in machine learning and data mining. However, in large-scale real-world scenarios, the exact similarity computation has become daunting due to “3V” nature (volume, velocity and variety) of big data. In this case, the hashing techniques have been verified to efficiently conduct similarity estimation in terms of both theory and practice. Currently, MinHash is a popular technique for efficiently estimating the Jaccard similarity of binary sets and furthermore, weighted MinHash is generalized to estimate the generalized Jaccard similarity of weighted sets. This review focuses on categorizing and discussing the existing works of weighted MinHash algorithms. In this review, we mainly categorize the weighted MinHash algorithms into quantization-based approaches, “active index”-based ones and others, and show the evolution and inherent connection of the weighted MinHash algorithms, from the integer weighted MinHash ones to the real-valued weighted MinHash ones. Also, we have developed a Python toolbox for the algorithms, and released it in our github. We experimentally conduct a comprehensive study of the standard MinHash algorithm and the weighted MinHash ones in the similarity estimation error and the information retrieval task.

Abstract reasoning ability is fundamental to human intelligence. It enables humans to uncover relations among abstract concepts and further deduce implicit rules from the relations. As a well-known abstract visual reasoning task, Raven's Progressive Matrices (RPM) are widely used in human IQ tests. Although extensive research has been conducted on RPM solvers with machine intelligence, few studies have considered further advancing the standard answer-selection (classification) problem to a more challenging answer-painting (generating) problem, which can verify whether the model has indeed understood the implicit rules. In this paper we aim to solve the latter one by proposing a deep latent variable model, in which multiple Gaussian processes are employed as priors of latent variables to separately learn underlying abstract concepts from RPMs; thus the proposed model is interpretable in terms of concept-specific latent variables. The latent Gaussian process also provides an effective way of extrapolation for answer painting based on the learned concept-changing rules. We evaluate the proposed model on RPM-like datasets with multiple continuously-changing visual concepts. Experimental results demonstrate that our model requires only few training samples to paint high-quality answers, generate novel RPM panels, and achieve interpretability through concept-specific latent variables.

We present a framework called Acquired Deep Impressions (ADI) which continuously learns knowledge of objects as ``impressions'' for compositional scene understanding. In this framework, the model first acquires knowledge from scene images containing a single object in a supervised manner, and then continues to learn from novel multi-object scene images which may contain objects that have not been seen before without any further supervision, under the guidance of the learned knowledge as humans do. By memorizing impressions of objects into parameters of neural networks and applying the generative replay strategy, the learned knowledge can be reused to generate images with pseudo-annotations and in turn assist the learning of novel scenes. The proposed ADI framework focuses on the acquisition and utilization of knowledge, and is complementary to existing deep generative models proposed for compositional scene representation. We adapt a base model to make it fall within the ADI framework and conduct experiments on two types of datasets. Empirical results suggest that the proposed framework is able to effectively utilize the acquired impressions and improve the scene decomposition performance.

Image super-resolution, which is often regarded as a preprocessing procedure of scene text recognition, aims to recover the realistic features from a low-resolution text image. It has always been challenging due to large variations in text shapes, fonts, backgrounds, etc. However, most existing methods employ generic super-resolution frameworks to handle scene text images while ignoring text-specific properties such as text-level layouts and character-level details. In this paper, we establish a text-focused super-resolution framework, called Scene Text Telescope (STT). In terms of text-level layouts, we propose a Transformer-Based Super-Resolution Network (TBSRN) containing a Self-Attention Module to extract sequential information, which is robust to tackle the texts in arbitrary orientations. In terms of character-level details, we propose a Position-Aware Module and a Content-Aware Module to highlight the position and the content of each character. By observing that some characters look indistinguishable in low-resolution conditions, we use a weighted cross-entropy loss to tackle this problem. We conduct extensive experiments, including text recognition with pre-trained recognizers and image quality evaluation, on TextZoom and several scene text recognition benchmarks to assess the super-resolution images. The experimental results show that our STT can indeed generate text-focused super-resolution images and outperform the existing methods in terms of recognition accuracy.

The Dirichlet Belief Network (DirBN) was recently proposed as a promising deep generative model to learn interpretable deep latent distributions for objects. However, its current representation capability is limited since its latent distributions across different layers is prone to form similar patterns and can thus hardly use multi-layer structure to form flexible distributions. In this work, we propose Poisson-randomised Dirichlet Belief Networks (Pois-DirBN), which allows large mutations for the latent distributions across layers to enlarge the representation capability. Based on our key idea of inserting Poisson random variables in the layer-wise connection, Pois-DirBN first introduces a component-wise propagation mechanism to enable latent distributions to have large variations across different layers. Then, we develop a layer-wise Gibbs sampling algorithm to infer the latent distributions, leading to a larger number of effective layers compared to DirBN. In addition, we integrate out latent distributions and form a multi-stochastic deep integer network, which provides an alternative view on Pois-DirBN. We apply Pois-DirBN to relational modelling and validate its effectiveness through improved link prediction performance and more interpretable latent distribution visualisations. The code can be downloaded at https://github.com/xuhuifan/Pois_DirBN.

Chinese character recognition has attracted much research interest due to its wide applications. Although it has been studied for many years, some issues in this field have not been completely resolved yet, \textit{e.g.} the zero-shot problem. Previous character-based and radical-based methods have not fundamentally addressed the zero-shot problem since some characters or radicals in test sets may not appear in training sets under a data-hungry condition. Inspired by the fact that humans can generalize to know how to write characters unseen before if they have learned stroke orders of some characters, we propose a stroke-based method by decomposing each character into a sequence of strokes, which are the most basic units of Chinese characters. However, we observe that there is a one-to-many relationship between stroke sequences and Chinese characters. To tackle this challenge, we employ a matching-based strategy to transform the predicted stroke sequence to a specific character. We evaluate the proposed method on handwritten characters, printed artistic characters, and scene characters. The experimental results validate that the proposed method outperforms existing methods on both character zero-shot and radical zero-shot tasks. Moreover, the proposed method can be easily generalized to other languages whose characters can be decomposed into strokes.

Bayesian nonparametric space partition (BNSP) models provide a variety of strategies for partitioning a D-dimensional space into a set of blocks, such that the data within the same block share certain kinds of homogeneity. BNSP models are applicable to many areas, including regression/classification trees, random feature construction, and relational modelling. This survey provides the first comprehensive review of this subject. We explore the current progress of BNSP research through three perspectives: (1) Partition strategies, where we review the various techniques for generating partitions and discuss their theoretical foundation, `self-consistency'; (2) Applications, where we detail the current mainstream usages of BNSP models and identify some potential future applications; and (3) Challenges, where we discuss current unsolved problems and possible avenues for future research.

The mutually-exciting Hawkes process (ME-HP) is a natural choice to model reciprocity, which is an important attribute of continuous-time edge (dyadic) data. However, existing ways of implementing the ME-HP for such data are either inflexible, as the exogenous (background) rate functions are typically constant and the endogenous (excitation) rate functions are specified parametrically, or inefficient, as inference usually relies on Markov chain Monte Carlo methods with high computational costs. To address these limitations, we discuss various approaches to model design, and develop three variants of non-parametric point processes for continuous-time edge modelling (CTEM). The resulting models are highly adaptable as they generate intensity functions through sigmoidal Gaussian processes, and so provide greater modelling flexibility than parametric forms. The models are implemented via a fast variational inference method enabled by a novel edge modelling construction. The superior performance of the proposed CTEM models is demonstrated through extensive experimental evaluations on four real-world continuous-time edge data sets.

Networks are ubiquitous in the real world. Link prediction, as one of the key problems for network-structured data, aims to predict whether there exists a link between two nodes. The traditional approaches are based on the explicit similarity computation between the compact node representation by embedding each node into a low-dimensional space. In order to efficiently handle the intensive similarity computation in link prediction, the hashing technique has been successfully used to produce the node representation in the Hamming space. However, the hashing-based link prediction algorithms face accuracy loss from the randomized hashing techniques or inefficiency from the learning to hash techniques in the embedding process. Currently, the Graph Neural Network (GNN) framework has been widely applied to the graph-related tasks in an end-to-end manner, but it commonly requires substantial computational resources and memory costs due to massive parameter learning, which makes the GNN-based algorithms impractical without the help of a powerful workhorse. In this paper, we propose a simple and effective model called #GNN, which balances the trade-off between accuracy and efficiency. #GNN is able to efficiently acquire node representation in the Hamming space for link prediction by exploiting the randomized hashing technique to implement message passing and capture high-order proximity in the GNN framework. Furthermore, we characterize the discriminative power of #GNN in probability. The extensive experimental results demonstrate that the proposed #GNN algorithm achieves accuracy comparable to the learning-based algorithms and outperforms the randomized algorithm, while running significantly faster than the learning-based algorithms. Also, the proposed algorithm shows excellent scalability on a large-scale network with the limited resources.

The Binary Space Partitioning-Tree (BSP-Tree) process was recently proposed as an efficient strategy for space partitioning tasks. Because it uses more than one dimension to partition the space, the BSP-Tree process is more efficient and flexible than conventional axis-aligned cut strategies. However, due to its batch learning setting, it is not well suited to large-scale classification and regression problems. In this paper, we develop an online BSP-Forest framework to address this limitation. With the arrival of new data, the resulting online algorithm can simultaneously expand the space coverage and refine the partition structure, with guaranteed universal consistency for classification problems. The effectiveness and competitive performance of the online BSP-Forest is verified via simulations.

The Dirichlet Belief Network~(DirBN) has been recently proposed as a promising approach in learning interpretable deep latent representations for objects. In this work, we leverage its interpretable modelling architecture and propose a deep dynamic probabilistic framework -- the Recurrent Dirichlet Belief Network~(Recurrent-DBN) -- to study interpretable hidden structures from dynamic relational data. The proposed Recurrent-DBN has the following merits: (1) it infers interpretable and organised hierarchical latent structures for objects within and across time steps; (2) it enables recurrent long-term temporal dependence modelling, which outperforms the one-order Markov descriptions in most of the dynamic probabilistic frameworks; (3) the computational cost scales to the number of positive links only. In addition, we develop a new inference strategy, which first upward-and-backward propagates latent counts and then downward-and-forward samples variables, to enable efficient Gibbs sampling for the Recurrent-DBN. We apply the Recurrent-DBN to dynamic relational data problems. The extensive experiment results on real-world data validate the advantages of the Recurrent-DBN over the state-of-the-art models in interpretable latent structure discovery and improved link prediction performance.

Humans perceive the seemingly chaotic world in a structured and compositional way with the prerequisite of being able to segregate conceptual entities from the complex visual scenes. The mechanism of grouping basic visual elements of scenes into conceptual entities is termed as perceptual grouping. In this work, we propose a new type of spatial mixture models with learnable priors for perceptual grouping. Different from existing methods, the proposed method disentangles the representation of an object into “shape” and “appearance” which are modeled separately by the mixture weights and the conditional probability distributions. More specifically, each object in the visual scene is modeled by one mixture component, whose mixture weights and the parameter of the conditional probability distribution are generated by two neural networks, respectively. The mixture weights focus on modeling spatial dependencies (i.e., shape) and the conditional probability distributions deal with intra-object variations (i.e., appearance). In addition, the background is separately modeled as a special component complementary to the foreground objects. Our extensive empirical tests on two perceptual grouping datasets demonstrate that the proposed method outperforms the stateof-the-art methods under most experimental configurations. The learned conceptual entities are generalizable to novel visual scenes and insensitive to the diversity of objects.

The Binary Space Partitioning (BSP)-Tree process is proposed to produce flexible 2-D partition structures which are originally used as a Bayesian nonparametric prior for relational modelling. It can hardly be applied to other learning tasks such as regression trees because extending the BSP-Tree process to a higher dimensional space is nontrivial. This paper is the first attempt to extend the BSP-Tree process to a d-dimensional ($d>2$) space. We propose to generate a cutting hyperplane, which is assumed to be parallel to $d-2$ dimensions, to cut each node in the d-dimensional BSP-tree. By designing a subtle strategy to sample two free dimensions from d dimensions, the extended BSP-Tree process can inherit the essential self-consistency property from the original version. Based on the extended BSP-Tree process, an ensemble model, which is named the BSP-Forest, is further developed for regression tasks. Thanks to the retained self-consistency property, we can thus significantly reduce the geometric calculations in the inference stage. Compared to its counterpart, the Mondrian Forest, the BSP-Forest can achieve similar performance with fewer cuts due to its flexibility. The BSP-Forest also outperforms other (Bayesian) regression forests on a number of real-world data sets.

We present a deep generative model which explicitly models object occlusions for compositional scene representation. Latent representations of objects are disentangled into location, size, shape, and appearance, and the visual scene can be generated compositionally by integrating these representations and an infinite-dimensional binary vector indicating presences of objects in the scene. By training the model to learn spatial dependences of pixels in the unsupervised setting, the number of objects, pixel-level segregation of objects, and presences of objects in overlapping regions can be estimated through inference of latent variables. Extensive experiments conducted on a series of specially designed datasets demonstrate that the proposed method outperforms two state-of-the-art methods when object occlusions exist.

Embeddings have become a key paradigm to learn graph representations and facilitate downstream graph analysis tasks. Existing graph embedding techniques either sample a large number of node pairs from a graph to learn node embeddings via stochastic optimization, or factorize a high-order proximity/adjacency matrix of the graph via expensive matrix factorization. However, these techniques usually require significant computational resources for the learning process, which hinders their applications on large-scale graphs. Moreover, the cosine similarity preserved by these techniques shows suboptimal efficiency in downstream graph analysis tasks, compared to Hamming similarity, for example. To address these issues, we propose NodeSketch, a highly-efficient graph embedding technique preserving high-order node proximity via recursive sketching. Specifically, built on top of an efficient data-independent hashing/sketching technique, NodeSketch generates node embeddings in Hamming space. For an input graph, it starts by sketching the self-loop-augmented adjacency matrix of the graph to output low-order node embeddings, and then recursively generates k-order node embeddings based on the self-loop-augmented adjacency matrix and (k-1)-order node embeddings. Our extensive evaluation compares NodeSketch against a sizable collection of state-of-the-art techniques using five real-world graphs on two graph analysis tasks. The results show that NodeSketch achieves state-of-the-art performance compared to these techniques, while showing significant speedup of 9x-372x in the embedding learning process and 1.19x-1.68x speedup when performing downstream graph analysis tasks.

In this work, we propose a probabilistic framework for relational data modelling and latent structure exploring. Given the possible feature information for the nodes in a network, our model builds up a deep architecture that can approximate to the possible nonlinear mappings between the nodes' feature information and latent representations. For each node, we incorporate all its neighborhoods' high-order structure information to generate latent representation, such that these latent representations are ``smooth'' in terms of the network. Since the latent representations are generated from Dirichlet distributions, we further develop a data augmentation trick to enable efficient Gibbs sampling for Ber-Poisson likelihood with Dirichlet random variables. Our model can be ready to apply to large sparse network as its computations cost scales to the number of positive links in the networks. The superior performance of our model is demonstrated through improved link prediction performance on a range of real-world datasets.

Min-Hash is a popular technique for efficiently estimating the Jaccard similarity of binary sets. Consistent Weighted Sampling (CWS) generalizes the Min-Hash scheme to sketch weighted sets and has drawn increasing interest from the community. Due to its constant-time complexity independent of the values of the weights, Improved CWS (ICWS) is considered as the state-of-the-art CWS algorithm. In this paper, we revisit ICWS and analyze its underlying mechanism to show that there actually exists dependence between the two components of the hash-code produced by ICWS, which violates the condition of independence. To remedy the problem, we propose an Improved ICWS (I2CWS) algorithm which not only shares the same theoretical computational complexity as ICWS but also abides by the required conditions of the CWS scheme. The experimental results on a number of synthetic data sets and real-world text data sets demonstrate that our I2CWS algorithm can estimate the Jaccard similarity more accurately, and also competes with or outperforms the compared methods, including ICWS, in classification and top-K retrieval, after relieving the underlying dependence.

Histogram-based similarity has been widely adopted in many machine learning tasks. However, measuring histogram similarity is a challenging task for streaming histograms, where the elements of a histogram are observed one after the other in an online manner. The ever-growing cardinality of histogram elements over the data streams makes any similarity computation inefficient in that case. To tackle this problem, we propose in this paper D2HistoSketch, a similarity-preserving sketching method for streaming histograms to efficiently approximate their Discriminative and Dynamic similarity. D2HistoSketch can fast and memory-efficiently maintain a set of compact and fixed-size sketches of streaming histograms to approximate the similarity between histograms. To provide high-quality similarity approximations, D2HistoSketch considers both discriminative and gradual forgetting weights for similarity measurement, and seamlessly incorporates them in the sketches. Based on both synthetic and real-world datasets, our empirical evaluation shows that our method is able to efficiently and effectively approximate the similarity between streaming histograms while outperforming state-of-the-art sketching methods. Compared to full streaming histograms with both discriminative and gradual forgetting weights in particular, D2HistoSketch is able to dramatically reduce the classification time (with a 7500x speedup) at the expense of a small loss in accuracy only (about 3.25 percent).

The Mondrian process represents an elegant and powerful approach for space partition modelling. However, as it restricts the partitions to be axis-aligned, its modelling flexibility is limited. In this work, we propose a self-consistent Binary Space Partitioning (BSP)-Tree process to generalize the Mondrian process. The BSP-Tree process is an almost surely right continuous Markov jump process that allows uniformly distributed oblique cuts in a two-dimensional convex polygon. The BSP-Tree process can also be extended using a non-uniform probability measure to generate direction differentiated cuts. The process is also self-consistent, maintaining distributional invariance under a restricted subdomain. We use Conditional-Sequential Monte Carlo for inference using the tree structure as the high-dimensional variable. The BSP-Tree process’s performance on synthetic data partitioning and relational modelling demonstrates clear inferential improvements over the standard Mondrian process and other related methods.

Attributed network embedding aims to learn a low-dimensional representation for each node of a network, considering both attributes and structure information of the node. However, the learning based methods usually involve substantial cost in time, which makes them impractical without the help of a powerful workhorse. In this paper, we propose a simple yet effective algorithm, named NetHash, to solve this problem only with moderate computing capacity. NetHash employs the randomized hashing technique to encode shallow trees, each of which is rooted at a node of the network. The main idea is to efficiently encode both attributes and structure information of each node by recursively sketching the corresponding rooted tree from bottom (i.e., the predefined highest-order neighboring nodes) to top (i.e., the root node), and particularly, to preserve as much information closer to the root node as possible. Our extensive experimental results show that the proposed algorithm, which does not need learning, runs significantly faster than the state-of-the-art learning-based network embedding methods while achieving competitive or even better performance in accuracy.

Stochastic partition models divide a multi-dimensional space into a number of rectangular regions, such that the data within each region exhibit certain types of homogeneity. Due to the nature of their partition strategy, existing partition models may create many unnecessary divisions in sparse regions when trying to describe data in dense regions. To avoid this problem we introduce a new parsimonious partition model -- the Rectangular Bounding Process (RBP) -- to efficiently partition multi-dimensional spaces, by employing a bounding strategy to enclose data points within rectangular bounding boxes. Unlike existing approaches, the RBP possesses several attractive theoretical properties that make it a powerful nonparametric partition prior on a hypercube. In particular, the RBP is self-consistent and as such can be directly extended from a finite hypercube to infinite (unbounded) space. We apply the RBP to regression trees and relational models as a flexible partition prior. The experimental results validate the merit of the RBP {in rich yet parsimonious expressiveness} compared to the state-of-the-art methods.

Many applications involve processing networked streaming data in a timely manner. Graph stream classification aims to learn a classification model from a stream of graphs with only one-pass of data, requiring real-time processing in training and prediction. This is a nontrivial task, as many existing methods require multipass of the graph stream to extract subgraph structures as features for graph classification which does not simultaneously satisfy "one-pass" and "real-time" requirements. In this paper, we propose an adaptive real-time graph stream classification method to address this challenge. We partition the unbounded graph stream data into consecutive graph chunks, each consisting of a fixed number of graphs and delivering a corresponding chunk-level classifier. We employ a random hashing function to compress the original node set of graphs in each chunk for fast feature detection when training chunk-level classifiers. Furthermore, a differential hashing strategy is applied to map unlimited increasing features (i.e., cliques) into a fixed-size feature space which is then used as a feature vector for stochastic learning. Finally, the chunk-level classifiers are weighted in an ensemble learning model for graph classification. The proposed method substantially speeds up the graph feature extraction and avoids unbounded graph feature growth. Moreover, it effectively offsets concept drifts in graph stream classification. Experiments on real-world and synthetic graph streams demonstrate that our method significantly outperforms existing methods in both classification accuracy and learning efficiency.

Existing graph classification usually relies on an exhaustive enumeration of substructure patterns, where the number of substructures expands exponentially w.r.t. with the size of the graph set. Recently, the Weisfeiler-Lehman (WL) graph kernel has achieved the best performance in terms of both accuracy and efficiency among state-of-the-art methods. However, it is still time-consuming, especially for large-scale graph classification tasks. In this paper, we present a K-Ary Tree based Hashing (KATH) algorithm, which is able to obtain competitive accuracy with a very fast runtime. The main idea of KATH is to construct a traversal table to quickly approximate the subtree patterns in WL using K-ary trees. Based on the traversal table, KATH employs a recursive indexing process that performs only r times of matrix indexing to generate all (r - 1)-depth K-ary trees, where the leaf node labels of a tree can uniquely specify the pattern. After that, the MinHash scheme is used to fingerprint the acquired subtree patterns for a graph. Our experimental results on both real world and synthetic data sets show that KATH runs significantly faster than state-of-the-art methods while achieving competitive or better accuracy.

Histogram-based similarity has been widely adopted in many machine learning tasks. However, measuring histogram similarity is a challenging task for streaming data, where the elements of a histogram are observed in a streaming manner. First, the ever-growing cardinality of histogram elements makes any similarity computation inefficient. Second, the concept-drift issue in the data streams also impairs the accurate assessment of the similarity. In this paper, we propose to overcome the above challenges with HistoSketch, a fast similarity-preserving sketching method for streaming histograms with concept drift. Specifically, HistoSketch is designed to incrementally maintain a set of compact and fixed-size sketches of streaming histograms to approximate similarity between the histograms, with the special consideration of gradually forgetting the outdated histogram elements. We evaluate HistoSketch on multiple classification tasks using both synthetic and real-world datasets. The results show that our method is able to efficiently approximate similarity for streaming histograms and quickly adapt to concept drift. Compared to full streaming histograms gradually forgetting the outdated histogram elements, HistoSketch is able to dramatically reduce the classification time (with a 7500x speedup) with only a modest loss in accuracy (about 3.5%).

Customer behavior modeling is important for businesses in order to understand, attract and retain customers. It is critical that the models are able to track the dynamics of customer behavior over time. We propose FC-CSM, a Customer Segmentation Model based on a Fragmentation-Coagulation process, which can track the evolution of customer segmentation, including the splitting and merging of customer groups. We conduct a case study using transaction data from a major Australian supermarket chain, where we: 1) show that our model achieves high fitness of purchase rate, outperforming models using mixture of Poisson processes; 2) compare the impact of promotions on customers for different products; and 3) track how customer groups evolve over time and how individual customers shift across groups. Our model provides valuable information to stakeholders about the different types of customers, how they change purchase behavior, and which customers are more receptive to promotion campaigns.

Min-Hash, which is widely used for efficiently estimating similarities of bag-of-words represented data, plays an increasingly important role in the era of big data. It has been extended to deal with real-value weighted sets -- Improved Consistent Weighted Sampling (ICWS) is considered as the state-of-the-art for this problem. In this paper, we propose a Practical CWS (PCWS) algorithm. We first transform the original form of ICWS into an equivalent expression, based on which we find some interesting properties that inspire us to make the ICWS algorithm simpler and more efficient in both space and time complexities. PCWS is not only mathematically equivalent to ICWS and preserves the same theoretical properties, but also saves 20% memory footprint and substantial computational cost compared to ICWS. The experimental results on a number of real-world text data sets demonstrate that PCWS obtains the same (even better) classification and retrieval performance as ICWS with 1/5~1/3 reduced empirical runtime.

Stochastic partition processes for exchangeable graphs produce axis-aligned blocks on a product space. In relational modeling, the resulting blocks uncover the underlying interactions between two sets of entities of the relational data. Although some flexible axis-aligned partition processes, such as the Mondrian process, have been able to capture complex interacting patterns in a hierarchical fashion, they are still in short of capturing dependence between dimensions. To overcome this limitation, we propose the Ostomachion process (OP), which relaxes the cutting direction by allowing for oblique cuts. The partitions generated by an OP are convex polygons that can capture inter-dimensional dependence. The OP also exhibits interesting properties: 1) Along the time line the cutting times can be characterized by a homogeneous Poisson process, and 2) on the partition space the areas of the resulting components comply with a Dirichlet distribution. We can thus control the expected number of cuts and the expected areas of components through hyper-parameters. We adapt the reversible-jump MCMC algorithm for inferring OP partition structures. The experimental results on relational modeling and decision tree classification have validated the merit of the OP.

The supermarkets often use sales promotions to attract customers and create brand loyalty. They would often like to know if their promotions are effective for various customers, so that better timing and more suitable rate can be planned in the future. Given a transaction data set collected by an Australian national supermarket chain, in this paper we conduct a case study aimed at discovering customers' long-term purchase patterns, which may be induced by preference changes, as well as short-term purchase patterns, which may be induced by promotions. Since purchase events of individual customers may be too sparse to model, we propose to discover a number of latent purchase patterns from the data. The latent purchase patterns are modeled via a mixture of non-homogeneous Poisson processes where each Poisson intensity function is composed by long-term and short-term components. Through the case study, 1) we validate that our model can accurately estimate the occurrences of purchase events; 2) we discover easy-to-interpret long-term gradual changes and short-term periodic changes in different customer groups; 3) we identify the customers who are receptive to promotions through the correlation between behavior patterns and the promotions, which is particularly worthwhile for target marketing.

Min-Hash, as a member of the Locality Sensitive Hashing (LSH) family for sketching sets, plays an important role in the big data era. It is widely used for efficiently estimating similarities of bag-of-words represented data and has been extended to dealing with multi-sets and real-value weighted sets. Improved Consistent Weighted Sampling (ICWS) has been recognized as the state-of-the-art for real-value weighted Min-Hash. However, the algorithmic implementation of ICWS is flawed because it violates the uniformity of the Min-Hash scheme. In this paper, we propose a Canonical Consistent Weighted Sampling (CCWS) algorithm, which not only retains the same theoretical complexity as ICWS but also strictly complies with the definition of Min-Hash. The experimental results demonstrate that the proposed CCWS algorithm runs faster than the state-of-the-arts while achieving similar classification performance on a number of real-world text data sets.

Water pipe failures can cause significant economic and social costs, hence have become the primary challenge to water utilities. In this paper, we propose a Bayesian nonparametric approach, namely the Dirichlet process mixture of hierarchical beta process model, for water pipe failure prediction. It can select high-risk pipes for physical condition assessment, thereby preventing disastrous failures proactively.

Stochastic partition processes in a product space play an important role in modeling relational data. Recent studies on the Mondrian process have introduced more flexibility into the block structure in relational models. A side-effect of such high flexibility is that, in data sparsity scenarios, the model is prone to overfit. In reality, relational entities are always associated with meta information, such as user profiles in a social network. In this paper, we propose a metadata dependent Mondrian process (MDMP) to incorporate meta information into the stochastic partition process in the product space and the entity allocation process on the resulting block structure. MDMP can not only encourage homogeneous relational interactions within blocks but also discourage meta-label diversity within blocks. Regularized by meta information, MDMP becomes more robust in data sparsity scenarios and easier to converge in posterior inference. We apply MDMP to link prediction and rating prediction and demonstrate that MDMP is more effective than the baseline models in prediction accuracy with a more parsimonious model structure.

Cross-domain collaborative filtering (CF) aims to share common rating knowledge across multiple related CF domains to boost the CF performance. In this paper, we view CF domains as a 2-D site-time coordinate system, on which multiple related domains, such as similar recommender sites or successive time-slices, can share group-level rating patterns. We propose a unified framework for cross-domain CF over the site-time coordinate system by sharing group-level rating patterns and imposing user/item dependence across domains. A generative model, say ratings over site-time (ROST), which can generate and predict ratings for multiple related CF domains, is developed as the basic model for the framework. We further introduce cross-domain user/item dependence into ROST and extend it to two real-world cross-domain CF scenarios: 1) ROST (sites) for alleviating rating sparsity in the target domain, where multiple similar sites are viewed as related CF domains and some items in the target domain depend on their correspondences in the related ones; and 2) ROST (time) for modeling user-interest drift over time, where a series of time-slices are viewed as related CF domains and a user at current time-slice depends on herself in the previous time-slice. All these ROST models are instances of the proposed unified framework. The experimental results show that ROST (sites) can effectively alleviate the sparsity problem to improve rating prediction performance and ROST (time) can clearly track and visualize user-interest drift over time.

Taxi service strategies, as the crowd intelligence of massive taxi drivers, are hidden in their historical time-stamped GPS traces. Mining GPS traces to understand the service strategies of skilled taxi drivers can benefit the drivers themselves, passengers, and city planners in a number of ways. This paper intends to uncover the efficient and inefficient taxi service strategies based on a large-scale GPS historical database of approximately 7600 taxis over one year in a city in China. First, we separate the GPS traces of individual taxi drivers and link them with the revenue generated. Second, we investigate the taxi service strategies from three perspectives, namely, passenger-searching strategies, passenger-delivery strategies, and service-region preference. Finally, we represent the taxi service strategies with a feature matrix and evaluate the correlation between service strategies and revenue, informing which strategies are efficient or inefficient. We predict the revenue of taxi drivers based on their strategies and achieve a prediction residual as less as 2.35 RMB/h,1which demonstrates that the extracted taxi service strategies with our proposed approach well characterize the driving behavior and performance of taxi drivers.

There have been a number of approximate algorithms for text similarity computation, such as min-wise hashing, random projection, and feature hashing, which are based on the bag-of-words representation. A limitation of their “flat-set” representation is that context information and semantic hierarchy cannot be preserved. In this paper, we aim to fast compute similarities between texts while also preserving context information. To take into account semantic hierarchy, we consider a notion of “multi-level exchangeability” which can be applied at word-level, sentence-level, paragraph-level, etc. We employ a nested-set to represent a multi-level exchangeable object. To fingerprint nested-sets for fast comparison, we propose a Recursive Min-wise Hashing (RMH) algorithm at the same computational cost of the standard min-wise hashing algorithm. Theoretical study and bound analysis confirm that RMH is a highly-concentrated estimator. The empirical studies show that the proposed context-preserving hashing method can significantly outperform min-wise hashing and feature hashing in accuracy at the same (or less) computational cost.

Traditional active learning methods require the labeler to provide a class label for each queried instance. The labelers are normally highly skilled domain experts to ensure the correctness of the provided labels, which in turn results in expensive labeling cost. To reduce labeling cost, an alternative solution is to allow nonexpert labelers to carry out the labeling task without explicitly telling the class label of each queried instance. In this paper, we propose a new active learning paradigm, in which a nonexpert labeler is only asked “whether a pair of instances belong to the same class”, namely, a pairwise label homogeneity. Under such circumstances, our active learning goal is twofold: (1) decide which pair of instances should be selected for query, and (2) how to make use of the pairwise homogeneity information to improve the active learner. To achieve the goal, we propose a “Pairwise Query on Max-flow Paths” strategy to query pairwise label homogeneity from a nonexpert labeler, whose query results are further used to dynamically update a Min-cut model (to differentiate instances in different classes). In addition, a “Confidence-based Data Selection” measure is used to evaluate data utility based on the Min-cut model’s prediction results. The selected instances, with inferred class labels, are included into the labeled set to form a closed-loop active learning process. Experimental results and comparisons with state-of-the-art methods demonstrate that our new active learning paradigm can result in good performance with nonexpert labelers.

The emergence of social tagging and crowdsourcing systems provides a unique platform where multiple weak labelers can form a crowd to fulfill a labeling task. Yet crowd labelers are often noisy, inaccurate, and have limited labeling knowledge, and worst of all, they act independently without seeking complementary knowledge from each other to improve labeling performance. In this paper, we propose a Self-Taught Active Learning (STAL) paradigm, where imperfect labelers are able to learn complementary knowledge from one another to expand their knowledge sets and benefit the underlying active learner. We employ a probabilistic model to characterize the knowledge of each labeler through which a weak labeler can learn complementary knowledge from a stronger peer. As a result, the self-taught active learning process eventually helps achieve high classification accuracy with minimized labeling costs and labeling errors.

Most studies on graph classification focus on designing fast and effective kernels. Several fast subtree kernels have achieved a linear time-complexity w.r.t. the number of edges under the condition that a common feature space (e.g., a subtree pattern list) is needed to represent all graphs. This will be infeasible when graphs are presented in a stream with rapidly emerging subtree patterns. In this case, computing a kernel matrix for graphs over the entire stream is difficult since the graphs in the expired chunks cannot be projected onto the unlimitedly expanding feature space again. This leads to a big trouble for graph classification over streams -- Different portions of graphs have different feature spaces. In this paper, we aim to enable large-scale graph classification over streams using the classical ensemble learning framework, which requires the data in different chunks to be in the same feature space. To this end, we propose a Nested Subtree Hashing (NSH) algorithm to recursively project the multi-resolution subtree patterns of different chunks onto a set of common low-dimensional feature spaces. We theoretically analyze the derived NSH kernel and obtain a number of favorable properties: 1) The NSH kernel is an unbiased and highly concentrated estimator of the fast subtree kernel. 2) The bound of convergence rate tends to be tighter as the NSH algorithm steps into a higher resolution. 3) The NSH kernel is robust in tolerating concept drift between chunks over a stream. We also empirically test the NSH kernel on both a large-scale synthetic graph data set and a real-world chemical compounds data set for anticancer activity prediction. The experimental results validate that the NSH kernel is indeed efficient and robust for graph classification over streams.

More and more location-based services become relying on the logical notion of a physical location, known as logical location (e.g. Starbucks, KFC). In this paper, we propose a new way to identify logical location using (1) a GPS-enabled mobile phone and (2) a wearable camera embedded in user's glasses. When a user with a wearable camera is detected paying attention to a certain physical location, all the logical locations within the error range of the GPS coordinates are considered as the matched candidates. We select the representative frames in the video stream corresponding to user's interested location in real-time and use multi-view images taken beforehand to represent each logical location. We then extract the Scale Invariant Feature Transform visual features from both the representative video frames and pre-stored images of candidate logical locations for video-image matching, the logical location that the user pays attention to can thus be identified. In order to differentiate the cases where users watch certain objects rather than a logical location in the street, we use Support Vector Machine to classify the two cases so that only the valid logical location is identified. Our proposed approach is proved weather and user independent, and it does not request additional user efforts compared with previous solutions. The results tested using a real-world dataset can achieve an average accuracy of 91.08%.

In real-world recommender systems, some users are easily influenced by new products and whereas others are unwilling to change their minds. So the preference varying speeds for users are different. Based on this observation, we propose a dynamic nonlinear matrix factorization model for collaborative filtering, aimed to improve the rating prediction performance as well as track the preference varying speeds for different users. We assume that user-preference changes smoothly over time, and the preference varying speeds for users are different. These two assumptions are incorporated into the proposed model as prior knowledge on user feature vectors, which can be learned efficiently by MAP estimation. The experimental results show that our method not only achieves state-of-the-art performance in the rating prediction task, but also provides an effective way to track user-preference varying speed.

Traditional active learning methods request experts to provide ground truths to the queried instances, which can be expensive in practice. An alternative solution is to ask nonexpert labelers to do such labeling work, which can not tell the definite class labels. In this paper, we propose a new active learning paradigm, in which a nonexpert labeler is only asked "whether a pair of instances belong to the same class". To instantiate the proposed paradigm, we adopt the MinCut algorithm as the base classifier. We first construct a graph based on the pairwise distance of all the labeled and unlabeled instances and then repeatedly update the unlabeled edge weights on the max-flow paths in the graph. Finally, we select an unlabeled subset of nodes with the highest prediction confidence as the labeled data, which are included into the labeled data set to learn a new classifier for the next round of active learning. The experimental results and comparisons, with state-of-the-art methods, demonstrate that our active learning paradigm can result in good performance with nonexpert labelers.

Cross-domain collaborative filtering (CF) is an emerging research topic in recommender systems. It aims to alleviate the sparsity problem in individual CF domains by transferring knowledge among related domains. In this paper, we will give a brief survey of the pilot studies in this research line in two dimensions: Collaborative Filtering Domains and Knowledge Transfer Styles. Some possible extensions for cross-domain CF will be discussed in the end.

Collaborative filtering (CF) techniques recommend items to users based on their historical ratings. In real-world scenarios, user interests may drift over time since they are affected by moods, contexts, and pop culture trends. This leads to the fact that a user's historical ratings comprise many aspects of user interests spanning a long time period. However, at a certain time slice, one user's interest may only focus on one or a couple of aspects. Thus, CF techniques based on the entire historical ratings may recommend inappropriate items. In this paper, we consider modeling user-interest drift over time based on the assumption that each user has multiple counterparts over temporal domains and successive counterparts are closely related. We adopt the cross-domain CF framework to share the static group-level rating matrix across temporal domains, and let user-interest distribution over item groups drift slightly between successive temporal domains. The derived method is based on a Bayesian latent factor model which can be inferred using Gibbs sampling. Our experimental results show that our method can achieve state-of-the-art recommendation performance as well as explicitly track and visualize user-interest drift over time.

Cross-domain collaborative filtering solves the sparsity problem by transferring rating knowledge across multiple domains. In this paper, we propose a rating-matrix generative model (RMGM) for effective cross-domain collaborative filtering. We first show that the relatedness across multiple rating matrices can be established by finding a shared implicit cluster-level rating matrix, which is next extended to a cluster-level rating model. Consequently, a rating matrix of any related task can be viewed as drawing a set of users and items from a user-item joint mixture model as well as drawing the corresponding ratings from the cluster-level rating model. The combination of these two models gives the RMGM, which can be used to fill the missing ratings for both existing and new users. A major advantage of RMGM is that it can share the knowledge by pooling the rating data from multiple tasks even when the users and items of these tasks do not overlap. We evaluate the RMGM empirically on three real-world collaborative filtering data sets to show that RMGM can outperform the individual models trained separately.

Scene is an important semantic unit for video analysis, retrieval and browsing. However, due to the lack of a generic algorithm, many studies focus on specific methods for certain video genes, e.g., news, sports, etc. In this paper, we propose a general framework for scene segmentation. First, we construct a graph, in which the elements encode the shot-to-shot coherent characteristics of a video clip based on visual similarity and temporal relation between shots. In this step, we only exploit the inherent property of video itself and it is independent of video genres. Second, spectral method is applied on the graph to group shots into scenes. The proposed method is not only simple but also effective to deal with organized features. Experimental results validate the robustness of our method on different kinds of videos.

For large-scale classification problems, the training samples can be clustered beforehand as a downsampling pre-process, and then only the obtained clusters are used for training. Motivated by such assumption, we proposed a classification algorithm, Support Cluster Machine (SCM), within the learning framework introduced by Vapnik. For the SCM, a compatible kernel is adopted such that a similarity measure can be handled not only between clusters in the training phase but also between a cluster and a vector in the testing phase. We also proved that the SCM is a general extension of the SVM with the RBF kernel. The experimental results confirm that the SCM is very effective for largescale classification problems due to significantly reduced computational costs for both training and testing and comparable classification accuracies. As a by-product, it provides a promising approach to dealing with privacy-preserving data mining problems.

Recent advantages in digital images and networks have made pornographic images more accessible than ever. Methods of detecting pornographic images have been proposed in some existing work. However, the skin detection modules adopted are all pixel-based and skin region shape features are rarely properly used. This paper presents a framework for pornographic image detection based on skin region information. Different from traditional works, our approach extracts color and texture features from arbitrary-shaped segmented regions. Then Gaussian mixture models are built for skin and non-skin region classification, and the skin map is produced based on the classification result. Finally, eigenregion features are used to describe the layout of skin regions on the whole image and pornographic images are detected according to the skin modality