Spatial multi-modal omics technologies have transformed biological research by enabling the simultaneous profiling of gene expression, protein abundance, and chromatin accessibility within their native spatial contexts. Despite these advances, accurately clustering rare cell types remains a major challenge due to data sparsity, high dimensionality, and limited annotated samples. While Graph Neural Networks (GNNs) have shown potential in modeling spatial omics data, their effectiveness is often constrained by the use of fixed K-nearest neighbor (KNN) graph structures, which fail to capture latent semantic relationships masked by sequencing noise. To overcome these limitations, we propose CRCT (Clustering Rare Cell Types): a novel framework that combines Implicit Semantic Data Augmentation (ISDA) with adaptive graph learning for spatial multi-modal omics analysis. Unlike traditional augmentation strategies that generate explicit synthetic samples, CRCT operates in the deep feature space by dynamically estimating intra-class covariance matrices and implicitly perturbing features along semantically meaningful directions. This enables effective augmentation for rare cell populations while preserving biological fidelity. Extensive experiments across four real-world datasets (HLN, MB, Stereo‑CITE‑seq, and SPOTS) and one synthetic benchmark demonstrate the state-of-the-art performance of CRCT, achieving improvements of up to +1.7 NMI and +7.8 ARI over strong baseline methods.

Enabling multi-task adaptation in pre-trained Low-Rank Adaptation (LoRA) models is crucial for enhancing their generalization capabilities. Most existing pre-trained LoRA fusion methods decompose weight matrices, sharing similar parameters, while fusion divergent ones. However, this paradigm inevitably induces inter-weight conflicts and leads to catastrophic domain forgetting. While incremental learning enables adaptation to multiple tasks, it struggles to achieve generalization in few-shot scenarios. Consequently, when the weight data follows a long-tailed distribution, it can lead to forgetting in the fused weights. To address this issue, we propose In-Context Meta LoRA Fusion (ICM-Fusion), a novel framework that synergizes meta-learning with in-context adaptation. The key innovation lies in our task vector arithmetic, which dynamically balances conflicting optimization directions across domains through learned manifold projections. ICM-Fusion obtains the optimal task vector orientation for the fused model in the latent space by adjusting the orientation of the task vectors. Subsequently, the fused LoRA is reconstructed by a self-designed Fusion VAE (F-VAE) to realize multi-task LoRA generation. We have conducted extensive experiments on visual and linguistic tasks, and the experimental results demonstrate that ICM-Fusion can be adapted to a wide range of architectural models and applied to various tasks. Compared to the current pre-trained LoRA fusion method, ICM-Fusion fused LoRA can significantly reduce the multi-tasking loss and can even achieve task enhancement in few-shot scenarios.

Diffusion models have been widely adopted in image and video generation. However, their complex network architecture leads to high inference overhead for its generation process. Existing diffusion quantization methods primarily focus on the quantization of the model structure while ignoring the impact of time-steps variation during sampling. At the same time, most current approaches fail to account for significant activations that cannot be eliminated, resulting in substantial performance degradation after quantization. To address these issues, we propose Time-Rotation Diffusion Quantization (TR-DQ), a novel quantization method incorporating time-step and rotation-based optimization. TR-DQ first divides the sampling process based on time-steps and applies a rotation matrix to smooth activations and weights dynamically. For different time-steps, a dedicated hyperparameter is introduced for adaptive timing modeling, which enables dynamic quantization across different time steps. Additionally, we also explore the compression potential of Classifier-Free Guidance (CFG-wise) to establish a foundation for subsequent work. TR-DQ achieves state-of-the-art (SOTA) performance on image generation and video generation tasks and a 1.38-1.89× speedup and 1.97-2.58× memory reduction in inference compared to existing quantization methods.

Zero-shot skeleton action recognition endeavors to classify novel action categories by transferring previously learned seen skeleton-semantic priors to unseen categories. However, current methods struggle to distinguish highly similar action categories, primarily due to the coarse-grained cross-modal alignment and non-discriminative representation space. To address these issues, we proposeSTAR++, a novel framework that aligns skeleton and semantics in a fine-grained and conditional manner. The key idea is to first establish region-level correspondences between body parts and semantic cues, and then utilize these local alignments to inform a global alignment process. This design is inspired by human visual cognition, which first attends to crucial local details before perceiving the broader scene. Concretely, we refine both skeleton and semantic representations with a dual-prompt attention mechanism driven by the structural decomposition of the human body and side information generated by a large language model (LLM). This encourages skeleton representations to be more compact within each class and semantic embeddings to be more separable across classes, which helps resolve ambiguity between highly similar actions and provides better interpretability of how unseen actions are perceived. Furthermore, we construct a region-aware holistic fusion module that aggregates these fine-grained features into a unified representation, yielding more discriminative holistic representations. Finally, the global alignment is conditioned on region-aware semantics feedback derived from fine-grained alignment, forming a conditional process that achieves more effective cross-modal alignment. Extensive experiments on four mainstream benchmarks demonstrate that our method achieves state-of-the-art performance in the zero-shot learning (ZSL) and generalized zero-shot learning (GZSL) settings.

The remarkable achievements of Large Language Models (LLMs) have captivated the attention of both academia and industry, transcending their initial role in dialogue generation. To expand the usage scenarios of LLM, some works enhance the effectiveness and capabilities of the model by introducing more external information, which is called the agent paradigm. Based on this idea, we propose a new method that integrates the agent paradigm into out-of-distribution (OOD) detection task, aiming to improve its robustness and adaptability. Our proposed method, Concept Matching with Agent (CMA), employs neutral prompts as agents to augment the CLIP-based OOD detection process. These agents function as dynamic observers and communication hubs, interacting with both In-distribution (ID) labels and data inputs to form vector triangle relationships. This triangular framework offers a more nuanced approach than the traditional binary relationship, allowing for better separation and identification of ID and OOD inputs. Our extensive experimental results showcase the superior performance of CMA over both zero-shot and training-required methods in a diverse array of real-world scenarios.

This paper investigates a challenging problem of zero-shot learning in the multi-label scenario (MLZSL), wherein the model is trained to recognize multiple unseen classes within a sample (e.g., an image) based on seen classes and auxiliary knowledge, e.g., semantic information. Existing methods usually resort to analyzing the relationship of various seen classes residing in a sample from the dimension of spatial or semantic characteristics and transferring the learned model to unseen ones. However, they neglect the integrity of local and global features. Although the use of the attention structure will accurately locate local features, especially objects, it will significantly lose its integrity, and the relationship between classes will also be affected. Rough processing of global features will also directly affect comprehensiveness. This neglect will make the model lose its grasp of the main components of the image. Relying only on the local existence of seen classes during the inference stage introduces unavoidable bias. In this paper, we propose a novel and comprehensive visual-semantic framework for MLZSL, dubbed Epsilon, to fully make use of such properties and enable a more accurate and robust visual-semantic projection. In terms of spatial information, we achieve effective refinement by group aggregating image features into several semantic prompts. It can aggregate semantic information rather than class information, preserving the correlation between semantics. In terms of global semantics, we use global forward propagation to collect as much information as possible to ensure that semantics are not omitted. Experiments on large-scale MLZSL benchmark datasets NUS-Wide and Open-Images-v4 demonstrate that the proposed Epsilon outperforms other state-of-the-art methods with large margins.

Recent years have witnessed the rise of Neural-enhanced Video Streaming (NeVS), which integrates neural restoration models into video codecs for higher compression-restoration performance. Despite its benefit, existing work has not well explored the full potential of NeVS paradigm, due to: (1) post-streaming restoration by decoder while lacking the proactive collaboration of encoder, (2) end-to-end optimization based on conventional rate-distortion theory, which has been verified that low distortion is not always a synonym for high perceptual quality, and (3) coupled design for domain-specific tasks that cannot generalize to various video codecs. Observing these limitations, our objective is not to incrementally present an improved restoration model. Instead, we focus on the encoder-decoder synergy, i.e., the codec, which is non-trivial since it inherently strikes the rate-distortion-perception trade-off of NeVS. Aiming at this target, we propose the Diffusion-enhanced Neural Codec (DeNC), a plug-and-play module for current NeVS paradigm, to significantly reduce the required bitrates while preserving high perceptual quality of restored videos. Our key design is twofold. First, DeNC improves the encoder's compression efficiency by simultaneously reducing the resolution and color bit-depth of frame referencing. Second, DeNC empowers the decoder with perception-oriented restoration capability by making its diffusion-based restoration process aware of the encoder's compression conditions. Real-world evaluations show that DeNC improves compression ratios with nearly an order of magnitude and achieves much higher restoration quality (e.g., 93+ VMAF and 23% higher MOS) over the latest baselines.

Zero-shot skeleton action recognition is a non-trivial task that requires robust unseen generalization with prior knowledge from only seen classes and shared semantics. Existing methods typically build the skeleton-semantics interactions by uncontrollable mappings and conspicuous representations, thereby can hardly capture the intricate and fine-grained relationship for effective cross-modal transferability. To address these issues, we propose a novel dyNamically Evolving dUal skeleton-semantic syneRgistic framework with the guidance of cOntext-aware side informatioN (dubbed Neuron), to explore more fine-grained cross-modal correspondence from micro to macro perspectives at both spatial and temporal levels, respectively. Concretely, 1) we first construct the spatial-temporal evolving micro-prototypes and integrate dynamic context-aware side information to capture the intricate and synergistic skeleton-semantic correlations step-by-step, progressively refining cross-model alignment; and 2) we introduce the spatial compression and temporal memory mechanisms to guide the growth of spatial-temporal micro-prototypes, enabling them to absorb structure-related spatial representations and regularity-dependent temporal patterns. Notably, such processes are analogous to the learning and growth of neurons, equipping the framework with the capacity to generalize to novel unseen action categories. Extensive experiments on various benchmark datasets demonstrated the superiority of the proposed method1.

Zero-shot learning (ZSL) aims to recognize unseen classes without labeled training examples by leveraging class-level semantic descriptors such as attributes. A fundamental challenge in ZSL is semantic misalignment, where semantic-unrelated information involved in visual features introduce ambiguity to visual-semantic interaction. Unlike existing methods that suppress semantic-unrelated information post hoc either in the feature space or the model space, we propose addressing this issue at the input stage, preventing semantic-unrelated patches from propagating through the network. To this end, we introduce Semantically contextualized VIsual Patches (SVIP) for ZSL, a transformer-based framework designed to enhance visual-semantic alignment. Specifically, we propose a self-supervised patch selection mechanism that preemptively learns to identify semantic-unrelated patches in the input space. This is trained with the supervision from aggregated attention scores across all transformer layers, which estimate each patch's semantic score. As removing semantic-unrelated patches from the input sequence may disrupt object structure, we replace them with learnable patch embeddings. With initialization from word embeddings, we can ensure they remain semantically meaningful throughout feature extraction. Extensive experiments on ZSL benchmarks demonstrate that SVIP achieves state-of-the-art performance results while providing more interpretable and semantically rich feature representations. Code is available at https://github.com/uqzhichen/SVIP.

Semi-supervised learning (SSL) has made notable advancements in medical image segmentation (MIS), particularly in scenarios with limited labeled data and significantly enhancing data utilization efficiency. Previous methods primarily focus on complex training strategies to utilize unlabeled data but neglect the importance of graph structural information. Different from existing methods, we propose a graph-based clustering for semi-supervised medical image segmentation (GraphCL) by jointly modeling graph data structure in a unified deep model. The proposed GraphCL model enjoys several advantages. Firstly, to the best of our knowledge, this is the first work to model the data structure information for semi-supervised medical image segmentation (SSMIS). Secondly, to get the clustered features across different graphs, we integrate both pairwise affinities between local image features and raw features as inputs. Extensive experimental results on three standard benchmarks show that the proposed GraphCL algorithm outperforms state-of-the-art semi-supervised medical image segmentation methods.

Compositional Zero-Shot Learning (CZSL) aims to recognize novel state-object compositions by leveraging the shared knowledge of their primitive components. Despite considerable progress, effectively calibrating the bias between semantically similar multimodal representations, as well as generalizing pre-trained knowledge to novel compositional contexts, remains an enduring challenge. In this paper, our interest is to revisit the conditional transport (CT) theory and its homology to the visual-semantics interaction in CZSL and further, propose a novel Trisets Consistency Alignment framework (dubbed TsCA) that well-addresses these issues. Concretely, we utilize three distinct yet semantically homologous sets, i.e., patches, primitives, and compositions, to construct pairwise CT costs to minimize their semantic discrepancies. To further ensure the consistency transfer within these sets, we implement a cycle-consistency constraint that refines the learning by guaranteeing the feature consistency of the self-mapping during transport flow, regardless of modality. Moreover, we extend the CT plans to an open-world setting, which enables the model to effectively filter out unfeasible pairs, thereby speeding up the inference as well as increasing the accuracy. Extensive experiments are conducted to verify the effectiveness of the proposed method. The code is available at https://github.com/keepgoingjkg/TsCA.

Domain adaptation (DA) is crucial for transferring models trained in one domain to perform well in a different, often unseen domain. Traditional methods, including unsupervised domain adaptation (UDA) and source-free domain adaptation (SFDA), have made significant progress. However, most existing DA methods rely heavily on Batch Normalization (BN) layers, which are not optimal in source-free settings, where the source domain is unavailable for comparison. In this study, we propose a novel method, ESBN, which addresses the challenge of domain shift by adjusting the placement of normalization layers and replacing BN with Batch-free Normalization (BFN). Unlike BN, BFN is less dependent on batch statistics and provides more robust feature representations through instance-specific statistics. We systematically investigate the effects of different BN layer placements across various network configurations and demonstrate that selective replacement with BFN improves generalization performance. Extensive experiments on multiple domain adaptation benchmarks show that our approach outperforms state-of-the-art methods, particularly in challenging scenarios such as Open-Partial Domain Adaptation (OPDA).

Conditional dependency present one of the trickiest problems in Compositional Zero-Shot Learning, leading to significant property variations of the same state (object) across different objects (states). To address this problem, existing approaches often adopt either all-to-one or one-to-one representation paradigms. However, these extremes create an imbalance in the seesaw between transferability and discriminability, favoring one at the expense of the other. Comparatively, humans are adept at analogizing and reasoning in a hierarchical clustering manner, intuitively grouping categories with similar properties to form cohesive concepts. Motivated by this, we propose Homogeneous Group Representation Learning (HGRL), a new perspective formulates state (object) representation learning as multiple homogeneous sub-group representation learning. HGRL seeks to achieve a balance between semantic transferability and discriminability by adaptively discovering and aggregating categories with shared properties, learning distributed group centers that retain group-specific discriminative features. Our method integrates three core components designed to simultaneously enhance both the visual and prompt representation capabilities of the model. Extensive experiments on three benchmark datasets validate the effectiveness of our method. Code is available at https://github.com/zjrao/HGRL.

Low-rank Adaptation (LoRA) has demonstrated remarkable capabilities for task specific fine-tuning. However, in scenarios that involve multiple tasks, training a separate LoRA model for each one results in considerable inefficiency in terms of storage and inference. Moreover, existing parameter generation methods fail to capture the correlations among these tasks, making multi-task LoRA parameter generation challenging. To address these limitations, we propose In-Context Meta LoRA (ICM-LoRA), a novel approach that efficiently achieves task-specific customization of large language models (LLMs). Specifically, we use training data from all tasks to train a tailored generator, Conditional Variational Autoencoder (CVAE). CVAE takes task descriptions as inputs and produces task-aware LoRA weights as outputs. These LoRA weights are then merged with LLMs to create task-specialized models without the need for additional fine-tuning. Furthermore, we utilize in-context meta-learning for knowledge enhancement and task mapping, to capture the relationship between tasks and parameter distributions. As a result, our method achieves more accurate LoRA parameter generation for diverse tasks using CVAE. ICM-LoRA enables more accurate LoRA parameter reconstruction than current parameter reconstruction methods and is useful for implementing task-specific enhancements of LoRA parameters. At the same time, our method occupies 283MB, only 1% storage compared with the original LoRA. The code is available at https://github.com/YihuaJerry/ICM-LoRA.

Personalized federated learning (PFL) tailors models to clients' unique data distributions while preserving privacy. However, existing aggregation-weight-based PFL methods often struggle with heterogeneous data, facing challenges in accuracy, computational efficiency, and communication overhead. We propose FedAPA, a novel PFL method featuring a server-side, gradient-based adaptive aggregation strategy to generate personalized models, by updating aggregation weights based on gradients of client-parameter changes with respect to the aggregation weights in a centralized manner. FedAPA guarantees theoretical convergence and achieves superior accuracy and computational efficiency compared to 10 PFL competitors across three datasets, with competitive communication overhead. The code and full proofs are available at: https://github.com/Yuxia-Sun/FL_FedAPA.

Exemplar-free class incremental learning (EF-CIL) is a nontrivial task that requires continuously enriching model capability with new classes while maintaining previously learned knowledge without storing and replaying any old class exemplars. An emerging theory-guided framework for CIL trains task-specific models for a shared network, shifting the pressure of forgetting to task-id prediction. In EF-CIL, task-id prediction is more challenging due to the lack of inter-task interaction (e.g., replays of exemplars). To address this issue, we conduct a theoretical analysis of the importance and feasibility of preserving a discriminative and consistent feature space, upon which we propose a novel method termed DCNet. Concretely, it progressively maps class representations into a hyperspherical space, in which different classes are orthogonally distributed to achieve ample inter-class separation. Meanwhile, it also introduces compensatory training to adaptively adjust supervision intensity, thereby aligning the degree of intra-class aggregation. Extensive experiments and theoretical analysis verified the superiority of DCNet. Code is available at https://github.com/Tianqi-Wang1/DCNet.

Click-through rate (CTR) prediction lies at the heart of the online advertising ecosystem and recommendation systems, helping to improve user engagement and platform revenue. With the advent of Pretrained Language Models (PLMs), researchers focus on incorporating textual features to enhance semantic understanding in CTR prediction. However, existing methods typically aggregate a wealth of textual features and encode the informative text into a single semantic embedding. This mechanism leads to entangled embedding that fails to capture fine-grained feature interactions, ultimately limiting CTR prediction performance. To address this issue, we propose Multi-faceted Semantic Disentanglement for CTR prediction (MSD-CTR), a novel framework designed to disentangle and leverage multi-faceted textual information. MSD-CTR consists of two key components: Disentangled Semantic Topic Model (DSTopic) and Topic Guided Disentangled Representation Learning (TopicDRL). DSTopic employs a disentangled generative process to extract multi-faceted knowledge from the entangled textual information. Meanwhile, TopicDRL integrates the extracted multi-faceted knowledge into CTR prediction and introduces two alignment losses to guide disentangled semantic embedding learning. Extensive experiments on four real-world datasets demonstrate that MSD-CTR outperforms existing CTR models, highlighting the effectiveness of disentangling textual information for better click-through rate prediction.

Recent advances in Generalized Zero-Shot Learning have demonstrated promising results by leveraging cross-modal attention mechanisms to model the nuanced relationships between semantic attributes and visual regions. However, we find that inherent attribute imbalance leads to significant attention disparities, i.e., those attributes with less sample weights often exhibit lower attention confidence and poorer localization accuracy. This attention illusion obviously leads to potential degraded or even counterproductive contributions to category determination. Conventional rebalancing approaches like resampling or reweighting are ineffective for attention calibration due to the complex interdependencies among attributes in one object and the absence of region-level annotation guidance. In this paper, we propose a novel coarse-to-fine framework termed Hierarchical Progressive Attention Network (HPAN), which leverages latent attention consistency across attributes to rectify the focus deviation of minor attributes. HPAN comprises two synergistic components, Super-Attribute Guided Generable Attention (SGGA) and Rebalanced Attribute Attention Calibration (RAAC). SGGA employs super-attributes to establish unified attention regions for both major and minor attributes, subsequently propagating these attention masks to RAAC for calibration. RAAC specializes in capturing fine-grained attribute-region interaction relations, with its output attention masks serving as supervisory signals to iteratively optimize the coarse attention of SGGA. Extensive experiments on three benchmark datasets verify the effectiveness of the proposed method. Code is available at github.com/zjrao/HPAN.

Unlabeled text representation learning (UTRL), encompassing static word embeddings such as Word2Vec and contextualized word embeddings such as bidirectional encoder representations from transformer (BERT), aims to capture semantic word relationships in a low-dimensional space without the need for manual labeling. These word embeddings are invaluable for downstream tasks such as document classification and clustering. However, the surge of short texts generated daily on social media platforms results in sparse word cooccurrences, compromising UTRL outcomes. Contextualized models such as recurrent neural network (RNN) and BERT, while impressive, often struggle with predicting the next word due to sparse word sequences in short texts. To address this, we introduce CGraphNet, a contrastive graph context prediction model designed for UTRL. This approach converts short texts into graphs, establishing links between sequentially occurring words. Information from the next word and its neighbors informs the target prediction, a process referred to as graph context prediction, mitigating sparse word cooccurrence issues in brief sentences. To minimize noise, an attention mechanism assigns importance to neighbors, while a contrastive objective encourages more distinctive representations by comparing the target word with its neighbors. Our experiments demonstrate CGraphNet's superior performance over other baselines, particularly in classification and clustering tasks on real-world datasets.

Node classification tasks aim to assign labels or categories to entire graphs based on their structural properties or node attributes. It can be adopted for various types of graph systems, including but not limited to network traffic, biological networks, knowledge graphs, etc., especially to social networks. This problem is well-studied, and solutions have demonstrated significant success in numerous real-world applications. However, in the situation where emerging categories are scarce or even have no labeled data, classical methods perform poorly on the whole, which has attracted growing attention. Based on this, in this article, we divide researches for node classification in social networks into two broad categories: traditional methods and novel strategies (few-shot/zero-shot learning). In traditional node classification methods, we summarize some classical methods for both homogeneous and heterogeneous networks, which includes unsupervised classifier, matrix factorization techniques, supervised methods, random-walk, and meta-path. Meanwhile, we introduce novel methods in few-shot or zero-shot learning. The article outlines the technical principles of various methods and analyzes their performance across different classes. It further summarizes the benchmark datasets used for evaluating node classification tasks. Finally, the major opportunities, challenges, and future research directions in few-shot and zero-shot learning for node classification in graph scenarios are discussed.

Personalized federated learning (pFL) is to collaboratively train non-identical machine learning models for different clients to adapt to their heterogeneously distributed datasets. State-of-the-art pFL approaches pay much attention on exploiting clients’ inter-similarities to facilitate the collaborative learning process, meanwhile, can barely escape from the irrelevant knowledge pooling that is inevitable during the aggregation phase, and thus hindering the optimization convergence and degrading the personalization performance. To tackle such conflicts between facilitating collaboration and promoting personalization, we propose a novel pFL framework, dubbed pFedC, to first decompose the global aggregated knowledge into several compositional branches, and then selectively reassemble the relevant branches for supporting conflicts-aware collaboration among contradictory clients. Specifically, by reconstructing each local model into a shared feature extractor and multiple decomposed task-specific classifiers, the training on each client transforms into a mutually reinforced and relatively independent multi-task learning process, which provides a new perspective for pFL. Besides, we conduct a purified knowledge aggregation mechanism via quantifying the combination weights for each client to capture clients’ common prior, as well as mitigate potential conflicts from the divergent knowledge caused by the heterogeneous data. Extensive experiments over various models and datasets demonstrate the effectiveness and superior performance of the proposed algorithm.

Skeleton-based action representation learning aims to interpret and understand human behaviors by encoding the skeleton sequences, which can be categorized into two primary training paradigms: supervised learning and self-supervised learning. However, the former one-hot classification requires labor-intensive predefined action categories annotations, while the latter involves skeleton transformations (e.g., cropping) in the pretext tasks that may impair the skeleton structure. To address these challenges, we introduce a novel skeleton-based training framework (C$^{2}$VL) based onCross-modalContrastive learning that uses the progressive distillation to learn task-agnostic human skeleton action representation from theVision-Language knowledge prompts. Specifically, we establish the vision-language action concept space through vision-language knowledge prompts generated by pre-trained large multimodal models (LMMs), which enrich the fine-grained details that the skeleton action space lacks. Moreover, we propose the intra-modal self-similarity and inter-modal cross-consistency softened targets in the cross-modal representation learning process to progressively control and guide the degree of pulling vision-language knowledge prompts and corresponding skeletons closer. These soft instance discrimination and self-knowledge distillation strategies contribute to the learning of better skeleton-based action representations from the noisy skeleton-vision-language pairs. During the inference phase, our method requires only the skeleton data as the input for action recognition and no longer for vision-language prompts. Extensive experiments on NTU RGB+D 60, NTU RGB+D 120, and PKU-MMD datasets demonstrate that our method outperforms the previous methods and achieves state-of-the-art results.

Malware open-set recognition (MOSR) is an emerging research domain that aims at jointly classifying malware samples from known families and detecting the ones from novel unknown families, respectively. Existing works mostly rely on a well-trained classifier considering the predicted probabilities of each known family with a threshold-based detection to achieve the MOSR. However, our observation reveals that the feature distributions of malware samples are extremely similar to each other even between known and unknown families. Thus, the obtained classifier may produce overly high probabilities of testing unknown samples toward known families and degrade the model performance. In this article, we propose the multi\modal dual-embedding networks, dubbed MDENet, to take advantage of comprehensive malware features from different modalities to enhance the diversity of malware feature space, which is more representative and discriminative for down-stream recognition. Concretely, we first generate a malware image for each observed sample based on their numeric features using our proposed numeric encoder with a re- designed multiscale CNN structure, which can better explore their statistical and spatial correlations. Besides, we propose to organize tokenized malware features into a sentence for each sample considering its behaviors and dynamics, and utilize language models as the textual encoder to transform it into a representable and computable textual vector. Such parallel multimodal encoders can fuse the above two components to enhance the feature diversity. Last, to further guarantee the open-set recognition (OSR), we dually embed the fused multimodal representation into one primary space and an associated sub-space, i.e., discriminative and exclusive spaces, with contrastive sampling and -bounded enclosing sphere regularizations, which resort to classification and detection, respectively. Moreover, we also enrich our previously proposed large-scaled malware dataset MAL-100 with multimodal characteristics and contribute an improved version dubbed MAL-100+. Experimental results on the widely used malware dataset Mailing and the proposed MAL-100+ demonstrate the effectiveness of our method.

This article provides a novel framework for single-domain generalized object detection (i.e., Single-DGOD), where we are interested in learning and maintaining the semantic structures of self-augmented compound cross-domain samples to enhance the model's generalization ability. Different from domain generalized object detection (DGOD) trained on multiple source domains, Single-DGOD is far more challenging to generalize well to multiple target domains with only one single source domain. Existing methods mostly adopt a similar treatment from DGOD to learn domain-invariant features by decoupling or compressing the semantic space. However, there may exist two potential limitations: 1) pseudo attribute-label correlation due to extremely scarce single-domain data and 2) the semantic structural information is usually ignored, i.e., we found the affinities of instance-level semantic relations in samples are crucial to model generalization. In this article, we introduce semantic reasoning with compound domains (SRCD) for Single-DGOD. Specifically, our SRCD contains two main components, namely, the texture-based self-augmentation (TBSA) module and the local-global semantic reasoning (LGSR) module. TBSA aims to eliminate the effects of irrelevant attributes associated with labels, such as light, shadow, and color, at the image level by a light-yet-efficient self-augmentation. Moreover, LGSR is used to further model the semantic relationships on instance features to uncover and maintain the intrinsic semantic structures. Extensive experiments on multiple benchmarks demonstrate the effectiveness of the proposed SRCD. Code is available at github.com/zjrao/SRCD.

Mixed-precision quantization mostly predetermines the model bit-width settings before actual training due to the non-differential bit-width sampling process, obtaining suboptimal performance. Worse still, the conventional static quality-consistent training setting, i.e., all data is assumed to be of the same quality across training and inference, overlooks data quality changes in real-world applications which may lead to poor robustness of the quantized models. In this article, we propose a novel data quality-aware mixed-precision quantization framework, dubbed DQMQ, to dynamically adapt quantization bit-widths to different data qualities. The adaption is based on a bit-width decision policy that can be learned jointly with the quantization training. Concretely, DQMQ is modeled as a hybrid reinforcement learning (RL) task that combines model-based policy optimization with supervised quantization training. By relaxing the discrete bit-width sampling to a continuous probability distribution that is encoded with few learnable parameters, DQMQ is differentiable and can be directly optimized end-to-end with a hybrid optimization target considering both task performance and quantization benefits. Trained on mixed-quality image datasets, DQMQ can implicitly select the most proper bit-width for each layer when facing uneven input qualities. Extensive experiments on various benchmark datasets and networks demonstrate the superiority of DQMQ against existing fixed/mixed-precision quantization methods.

Personalized recommendation is of paramount importance in online content platforms like Kuai and Tencent. To ensure accurate recommendations, it is crucial to consider multi-modal information in both items and user-user/item interactions. While existing works on multimedia recommendation have made strides in leveraging multi-modal contents to enrich item representations, many of them overlook the practical scenario of multiple modality missing. As a result, the performance of recommendation systems can be significantly compromised in such cases. In this paper, we introduce a novel multi-modal adversarial method called$MMAM$, which aims to provide reliable personalized recommendation services even in the presence of uncertain missing modalities. The core idea behind$MMAM$is to design a generator that can effectively encode both user-user/item interactions and multi-modal contents, taking into account various missing cases. The generator is trained to learn transferable features from different combinations of missing modalities in order to deceive a discriminative classifier. Additionally, we propose a modal discriminator that can classify the missing cases of multi-modalities, further enhancing the capability of the model. Moreover, a well-equipped predictor utilizes the transferable features to predict potential user interests. To improve the prediction accuracy, we design a type discriminator that enhances the classification of link types. By employing a mini-max game between the generator and the discriminators,$MMAM$successfully obtains transferable features that encompass multi-modal contents, even when facing uncertain missing modalities. We conduct extensive experiments on industrial datasets, including Kuai and Tencent. Comparing with state-of-the-art approaches, MMAM achieves improvements in personalized recommendation tasks under uncertain missing modalities. MMAM holds promise for enhancing multi-modal personalized recommendations in real-world applications.

Open-World Compositional Zero-shot Learning (OW-CZSL) aims to recognize novel compositions of state and object primitives in images with no priors on the compositional space, which induces a tremendously large output space containing all possible state-object compositions. Existing works either learn the joint compositional state-object embedding or predict simple primitives with separate classifiers. However, the former method heavily relies on external word embedding methods, and the latter ignores the interactions of interdependent primitives, respectively. In this paper, we revisit the primitive prediction approach and propose a novel method, termed Progressive Cross-primitive Compatibility (ProCC), to mimic the human learning process for OW-CZSL tasks. Specifically, the cross-primitive compatibility module explicitly learns to model the interactions of state and object features with the trainable memory units, which efficiently acquires cross-primitive visual attention to reason high-feasibility compositions, without the aid of external knowledge. Moreover, to alleviate the invalid cross-primitive interactions, especially for partial-supervision conditions (pCZSL), we design a progressive training paradigm to optimize the primitive classifiers conditioned on pre-trained features in an easy-to-hard manner. Extensive experiments on three widely used benchmark datasets demonstrate that our method outperforms other representative methods on both OW-CZSL and pCZSL settings by large margins.

This paper investigates a new, practical, but challenging problem named Non-exemplar Online Class-incremental continual Learning (NO-CL), which aims to preserve the discernibility of base classes without buffering data examples and efficiently learn novel classes continuously in a single-pass (i.e., online) data stream. The challenges of this task are mainly two-fold: (1) Both base and novel classes suffer from severe catastrophic forgetting as no previous samples are available for replay. (2) As the online data can only be observed once, there is no way to fully re-train the whole model, e.g., re-calibrate the decision boundaries via prototype alignment or feature distillation. In this paper, we propose a novel Dual-prototype Self-augment and Refinement method (DSR) for NO-CL problem, which consists of two strategies: 1) Dual class prototypes: vanilla and high-dimensional prototypes are exploited to utilize the pre-trained information and obtain robust quasi-orthogonal representations rather than example buffers for both privacy preservation and memory reduction. 2) Self-augment and refinement: Instead of updating the whole network, we optimize high-dimensional prototypes alternatively with the extra projection module based on self-augment vanilla prototypes, through a bi-level optimization problem. Extensive experiments demonstrate the effectiveness and superiority of the proposed DSR in NO-CL.

The explosive growth of video traffic on today's Internet promotes the rise of Neural-enhanced Video Streaming (NeVS), which effectively improves the rate-distortion trade-off by employing a cheap neural super-resolution model for quality enhancement on the receiver side. Missing by existing work, we reveal that the NeVS pipeline may suffer from a practical threat, where the crucial codec component (i.e., encoder for compression and decoder for restoration) can trigger adversarial attacks in a man-in-the-middle manner to significantly destroy video recovery performance and finally incurs the malfunction of downstream video perception tasks. In this paper, we are the first attempt to inspect the vulnerability of NeVS and discover a novel adversarial attack, called codec hijacking, where the injected invisible perturbation conspires with the malicious encoding matrix by reorganizing the spatial-temporal bit allocation within the bitstream size budget. Such a zero-day vulnerability makes our attack hard to defend because there is no visual distortion on the recovered videos until the attack happens. More seriously, this attack can be extended to diverse enhancement models, thus exposing a wide range of video perception tasks under threat. Evaluation based on state-of-the-art video codec benchmark illustrates that our attack significantly degrades the recovery performance of NeVS over previous attack methods. The damaged video quality finally leads to obvious malfunction of downstream tasks with over 75% success rate. We hope to arouse public attention on codec hijacking and its defence.

Federated learning (FL) has emerged as a powerful paradigm for learning from decentralized data, and federated domain generalization further considers the test dataset (target domain) is absent from the decentralized training data (source domains). However, most existing FL methods assume that domain labels are provided during training, and their evaluation imposes explicit constraints on the number of domains, which must strictly match the number of clients. Because of the underutilization of numerous edge devices and additional cross-client domain annotations in the real world, such restrictions may be impractical and involve potential privacy leaks. In this paper, we propose an efficient and novel approach, called Disentangled Prompt Tuning (DiPrompT), a method that tackles the above restrictions by learning adaptive prompts for domain generalization in a distributed manner. Specifically, we first design two types of prompts, i.e., global prompt to capture general knowledge across all clients and domain prompts to capture domain-specific knowledge. They eliminate the restriction on the one-to-one mapping between source domains and local clients. Furthermore, a dynamic query metric is introduced to automatically search the suitable domain label for each sample, which includes two-substep text-image alignments based on prompt tuning without labor-intensive annotation. Extensive experiments on multiple datasets demonstrate that our DiPrompT achieves superior domain generalization performance over state-of-the-art FL methods when domain labels are not provided, and even outperforms many centralized learning methods using domain labels.

Existing Knowledge Distillation (KD) methods typically focus on transferring knowledge from a large-capacity teacher to a low-capacity student model, achieving sub-stantial success in unimodal knowledge transfer. However, existing methods can hardly be extended to Cross-Modal Knowledge Distillation (CMKD), where the knowledge is transferred from a teacher modality to a different student modality, with inference only on the distilled student modality. We empirically reveal that the modality gap, i.e., modality imbalance and soft label misalignment, incurs the in-effectiveness of traditional KD in CMKD. As a solution, we propose a novel f_ustomized Crossmodal Knowledge Distillation (C2KD). Specifically, to alleviate the modality gap, the pre-trained teacher performs bidirectional distillation with the student to provide customized knowledge. The On-the-Fly Selection Distillationi OFSD) strategy is applied to selectively filter out the samples with misaligned soft labels, where we distill cross-modal knowledge from non-target classes to avoid the modality imbalance issue. To further provide receptive cross-modal knowledge, proxy student and teacher, inheriting unimodal and cross-modal knowledge, is formulated to progressively transfer cross-modal knowledge through bidirectional distillation. Experimental results on audio-visual, image-text, and RGB-depth datasets demonstrate that our method can effectively transfer knowledge across modalities, achieving superior performance against traditional KD by a large margin.

Personalized federated learning (PFL) has gained great success in tackling the scenarios where target datasets are heterogeneous across the local clients. However, the application of the existing PFL methods to real-world setting is hindered by the common assumption that the test data on each client is in-distribution (IND) with respect to its training data. Due to the bias of training dataset, the modern machine learning model prefers to rely on shortcut which can perform well on the training data but fail to generalize to the unseen test data that is out-of-distribution (OOD). This pervasive phenomenon is called shortcut learning and has attracted plentiful efforts in centralized situations. In PFL, the limited data diversity on federated clients makes mitigating shortcut and meanwhile preserving personalization knowledge rather difficult. In this paper, we analyse this challenging problem by formulating the structural causal models (SCMs) for heterogeneous federated clients. From the proposed SCMs, we derive two significant causal signatures which inspire a provable shortcut discovery and removal method under federated learning, namely FedSDR. Specifically, FedSDR is divided into two steps: 1) utilizing the available training data distributed among local clients to discover all the shortcut features in a collaborative manner. 2) developing the optimal personalized causally invariant predictor for each client by eliminating the discovered shortcut features. We provide theoretical analysis to prove that our method can draw complete shortcut features and produce the optimal personalized invariant predictor that can generalize to unseen OOD data on each client. The experimental results on diverse datasets validate the superiority of FedSDR over the state-of-the-art PFL methods on OOD generalization performance.

Exploiting invariant relations and mitigating spurious correlation (a.k.a., shortcut) between representation and target across varied data distributions can tackle the challenging out-of-distribution (OOD) generalization problem. In personalized federated learning (PFL), heterogeneous data distribution across local clients offers the inherent prerequisites to extract the invariant features that maintain invariant relation with target. Nevertheless, personalized features are closely entangled with spurious features in PFL since they exhibit similar variability across different clients, which makes preserving personalization knowledge and eliminating shortcuts two conflicting objectives in PFL. To address the above challenge, we analyse the heterogeneous data generation on local clients through the lens of structured causal model and propose a crucial causal signature which can distinguish personalized features from spurious features with global invariant features as the anchor. Then the causal signature is quantified as an information-theoretic constraint that facilitates the shortcut-averse personalized invariant learning on each client. Theoretical analysis demonstrates our method, FedPIN, can yield a tighter bound on generalization error than the prevalent PFL approaches when train-test distribution shift exists on clients. Moreover, we provide a theoretical guarantee on the convergence rate of FedPIN in this paper. The results of extensive experiments show that our method can achieve superior OOD generalization performance compared with the state-of-the-art competitors.

Skeleton-based zero-shot action recognition aims to recognize unknown human actions based on the learned priors of the known skeleton-based actions and a semantic descriptor space shared by both known and unknown categories. However, previous works mainly focus on establishing the bridges between the known skeleton representation space and semantic descriptions space at the coarse-grained level for recognizing unknown action categories, ignoring the fine-grained alignment of these two spaces, resulting in suboptimal performance in distinguishing high-similarity action categories. To address these challenges, we propose a novel method via Side information and dual-prompTs learning for skeleton-based zero-shot Action Recognition (STAR) at the fine-grained level. Specifically, 1) we decompose the skeleton into several parts based on its topology structure and introduce the side information concerning multi-part descriptions of human body movements for alignment between the skeleton and the semantic space at the fine-grained level; 2) we design the visual-attribute and semantic-part prompts to improve the intra-class compactness within the skeleton space and inter-class separability within the semantic space, respectively, to distinguish the high-similarity actions. Extensive experiments show that our method achieves state-of-the-art performance in ZSL and GZSL settings on NTU RGB+D, NTU RGB+D 120, and PKU-MMD datasets.

This paper provides an efficient training-free painterly image harmonization (PIH) method, dubbed FreePIH, that leverages only a pre-trained diffusion model to achieve state-of-the-art harmonization results. Unlike existing methods that require either training auxiliary networks or fine-tuning a large pre-trained backbone, or both, to harmonize a foreground object with a painterly-style background image, our FreePIH tames the denoising process as a plug-in module for foreground image style transfer. Specifically, we find that the very last few steps of the denoising (i.e., generation) process strongly correspond to the stylistic information of images, and based on this, we propose to augment the latent features of both the foreground and background images with Gaussians for a direct denoising-based harmonization. To guarantee the fidelity of the harmonized image, we make use of latent features to enforce the consistency of the content and stability of the foreground objects in the latent space, and meanwhile, aligning both fore-/back-grounds with the same style. Moreover, to accommodate the generation with more structural and textural details, we further integrate text prompts to attend to the latent features, hence improving the generation quality. Quantitative and qualitative evaluations on COCO and LAION 5B datasets demonstrate that our method can surpass representative baselines by large margins.

Recent studies reveal that even highly biased dense networks can contain an invariant substructure with superior out-of-distribution (OOD) generalization. While existing works commonly seek these substructures using global sparsity constraints, the uniform imposition of sparse penalties across samples with diverse levels of spurious contents renders such methods suboptimal. The precise adaptation of model sparsity, specifically tailored for spurious features, remains a significant challenge. Motivated by the insight that in-distribution (ID) data containing spurious features may exhibit lower experiential risk, we propose a novel Spurious Feature-targeted Pruning framework, dubbed SFP, to induce the authentic invariant substructures without referring to the above concerns. Specifically, SFP distinguishes spurious features within ID instances during training by a theoretically validated threshold. It then penalizes the corresponding feature projections onto the model space, steering the optimization towards subspaces spanned by those invariant factors. Moreover, we also conduct detailed theoretical analysis to provide a rationality guarantee and a proof framework for OOD structures based on model sparsity. Experiments on various OOD datasets show that SFP can significantly outperform both structure-based and non-structure-based OOD generalization state-of-the-art (SOTA) methods by large margins.

Recent research has underscored the efficacy of Graph Neural Networks (GNNs) in modeling diverse geometric structures within graph data. However, real-world graphs typically exhibit geometrically heterogeneous characteristics, rendering the confinement to a single geometric paradigm insufficient for capturing their intricate structural complexities. To address this limitation, we examine the performance of GNNs across various geometries through the lens of knowledge distillation (KD) and introduce a novel cross-geometric framework. This framework encodes graphs by integrating both Euclidean and hyperbolic geometries in a space-mixing fashion. Our approach employs multiple teacher models, each generating hint embeddings that encapsulate distinct geometric properties. We then implement a structure-wise knowledge transfer module that optimally leverages these embeddings within their respective geometric contexts, thereby enhancing the training efficacy of the student model. Additionally, our framework incorporates a geometric optimization network designed to bridge the distributional disparities among these embeddings. Experimental results demonstrate that our model-agnostic framework more effectively captures topological graph knowledge, resulting in superior performance of the student models when compared to traditional KD methodologies.

Real-time video perception tasks are often challenging on resource-constrained edge devices due to the issues of accuracy drop and hardware overhead, where saving computations is the key to performance improvement. Existing methods either rely on domain-specific neural chips or priorly searched models, which require specialized optimization according to different task properties. These limitations motivate us to design a general and task-independent methodology, called Patch Automatic Skip Scheme (PASS), which supports diverse video perception settings by decoupling acceleration and tasks. The gist is to capture inter-frame correlations and skip redundant computations at patch level, where the patch is a non-overlapping square block in visual. PASS equips each convolution layer with a learnable gate to selectively determine which patches could be safely skipped without degrading model accuracy. Specifically, we are the first to construct a self-supervisory procedure for gate optimization, which learns to extract contrastive representations from frame sequences. The pre-trained gates can serve as plug-and-play modules to implement patch-skippable neural backbones, and automatically generate proper skip strategy to accelerate different video-based downstream tasks, e.g., outperforming state-of-the-art MobileHumanPose in 3D pose estimation and FairMOT in multiple object tracking, by up to 9.43 × and 12.19 × speedups, respectively, on NVIDIA Jetson Nano devices.

Zero-shot learning (ZSL) is an extreme case of transfer learning that aims to recognize samples (e.g., images) of unseen classes relying on a train-set covering only seen classes and a set of auxiliary knowledge (e.g., semantic descriptors). Existing methods usually resort to constructing a visual-to-semantics mapping based on features extracted from each whole sample. However, since the visual and semantic spaces are inherently independent and may exist in different manifolds, these methods may easily suffer from the domain bias problem due to the knowledge transfer from seen to unseen classes. Unlike existing works, this paper investigates the fine-grained ZSL from a novel perspective of sample-level graph. Specifically, we decompose an input into several fine-grained elements and construct a graph structure per sample to measure and utilize element-granularity relations within each sample. Taking advantage of recently developed graph neural networks (GNNs), we formulate the ZSL problem to a graph-to-semantics mapping task, which can better exploit element-semantics correlation and local sub-structural information in samples. Experimental results on the widely used benchmark datasets demonstrate that the proposed method can mitigate the domain bias problem and achieve competitive performance against other representative methods.

Recent studies usually approach multi-label zeroshot learning (MLZSL) with visual-semantic mapping on spatial-class correlation, which can be computationally costly, and worse still, fails to capture fine-grained classspecific semantics. We observe that different channels may usually have different sensitivities on classes, which can correspond to specific semantics. Such an intrinsic channelclass correlation suggests a potential alternative for the more accurate and class-harmonious feature representations. In this paper, our interest is to fully explore the power of channel-class correlation as the unique base for MLZSL. Specifically, we propose a light yet efficient Multi-Label Multi-Layer Perceptron-based Encoder, dubbed (ML)2P-Encoder, to extract and preserve channel-wise semantics. We reorganize the generated feature maps into several groups, of which each of them can be trained independently with (ML)2P-Encoder. On top of that, a global groupwise attention module is further designed to build the multilabel specific class relationships among different classes, which eventually fulfills a novel Channel-Class Correlation MLZSL framework (C3-MLZSL)11Released code:github.com/simonzmliu/cvpr23_mlzsl. Extensive experiments on large-scale MLZSL benchmarks including NUS-WIDE and Open-Images-V4 demonstrate the superiority of our model against other representative state-of-the-art models.

Compositional Zero-Shot Learning (CZSL) aims to recognize novel concepts formed by known states and objects during training. Existing methods either learn the combined state-object representation, challenging the generalization of unseen compositions, or design two classifiers to identify state and object separately from image features, ignoring the intrinsic relationship between them. To jointly eliminate the above issues and construct a more robust CZSL system, we propose a novel framework termed Decomposed Fusion with Soft Prompt (DFSP)11Code is available at: https://github.corn/Forest-art/DFSP.git, by involving vision-language models (VLMs)for unseen composition recognition. Specifically, DFSP constructs a vector combination of learnable soft prompts with state and object to establish the joint representation of them. In addition, a cross-modal decomposed fusion module is designed between the language and image branches, which decomposes state and object among language features instead of image features. Notably, being fused with the decomposed features, the image features can be more expressive for learning the relationship with states and objects, respectively, to improve the response of unseen compositions in the pair space, hence narrowing the domain gap between seen and unseen sets. Experimental results on three challenging benchmarks demonstrate that our approach significantly outperforms other state-of-the-art methods by large margins.

Federated learning (FL) enables collaborative model training among distributed devices without data sharing, but existing FL suffers from poor scalability because of global model synchronization. To address this issue, hierarchical federated learning (HFL) has been recently proposed to let edge servers aggregate models of devices in proximity, while synchronizing via the cloud periodically. However, a critical open challenge about how to design a good synchronization scheme (when devices and edges should be synchronized) is still unsolved. Devices are heterogeneous in computing and communication capability, and their data could be non-IID. No existing work can well synchronize various roles (e.g., devices and edge) in HFL to guarantee high learning efficiency and accuracy. In this paper, we propose a learning-based synchronization scheme for HFL systems. By collecting data such as edge models, CPU usage, communication time, etc., we design a deep reinforcement learning-based approach to decide the frequencies of cloud aggregation and edge aggregation, respectively. The proposed scheme well considers device heterogeneity, non-IID data and device mobility, to maximize the training model accuracy while minimizing the energy overhead. We build an HFL testbed and conduct experiments using real data obtained from Raspberry Pi and Alibaba Cloud. Extensive experimental results have confirmed the effectiveness of Hwamei.

Federated learning (FL) is an emerging technique that trains massive and geographically distributed edge data while maintaining privacy. However, FL has inherent challenges in terms of fairness and computational efficiency due to the rising heterogeneity of edges, and thus usually results in sub-optimal performance in recent state-of-the-art (SOTA) solutions. In this paper, we propose a Customized Federated Learning (CFL) system to eliminate FL heterogeneity from multiple dimensions. Specifically, CFL tailors personalized models from the specially designed global model for each client jointly guided by an online trained model-search helper and a novel aggregation algorithm. Extensive experiments demonstrate that CFL has full-stack advantages for both FL training and edge reasoning and significantly improves the SOTA performance w.r.t. model accuracy (up to 7.2% in the non-heterogeneous environment and up to 21.8% in the heterogeneous environment), efficiency, and FL fairness.

Federated Distillation (FD) extends classic Federated Learning (FL) to a more general training framework that enables model-heterogeneous collaborative learning by Knowledge Distillation (KD) across multiple clients and the server. However, existing KD-based algorithms usually require a set of shared input samples for each client to produce soft-prediction for distillation. Worse still, such a manual selection is accompanied by careful deliberations or prior information on clients’ private data distribution, which is not in line with the privacy-preserving characteristic of classic FL. In this paper, we propose a novel training framework to achieve data-independent knowledge transfer by properly designing a distributed generative adversarial network (GAN) between the server and clients that can synthesize shared feature representations to facilitate the FD training. Specifically, we deploy a generator on the server and reuse each local model as a federated discriminator to form a lightweight efficient distributed GAN that can automatically synthesize simulated global feature representations for distillation. Moreover, since the synthesized feature representations are usually more faithful and homologous with global data distribution, faster and better training convergence can be obtained. Extensive experiments on different tasks and heterogeneous models demonstrate the effectiveness of the proposed framework on model accuracy and communication overhead.

Multi-label zero-shot learning extends conventional single-label zero-shot learning to a more realistic scenario that aims at recognizing multiple unseen labels of classes for each input sample. Existing works usually exploit attention mechanism to generate the correlation among different labels. However, most of them are usually biased on severalmajor classeswhile neglect most of theminor classeswith the same importance in input samples, and may thus result in overly diffused attention maps that cannot sufficiently coverminor classes. We argue that disregarding the connection between major and minor classes, i.e., correspond to the global and local information, respectively, is the cause of the problem. In this paper, we propose a novel framework of unbiased multi-label zero-shot learning, by considering various class-specific regions to calibrate the training process of the classifier. Specifically,Pyramid Feature Attention(PFA) is proposed to build the correlation between global and local information of samples to balance the presence of each class. Meanwhile, for the generated semantic representations of input samples, we proposeSemantic Attention(SA) to strengthen the element-wise correlation among these vectors, which can encourage the coordinated representation of them. Extensive experiments on the large-scale multi-label benchmarksMS-COCO,NUS-WIDEandOpen-Imagesdemonstrate that the proposed method surpasses other representative methods by significant margins.

We study the challenging task of malware recognition on both known and novel unknown malware families, called malware open-set recognition (MOSR). Previous works usually assume the malware families are known to the classifier in a close-set scenario, i.e., testing families are the subset or at most identical to training families. However, novel unknown malware families frequently emerge in real-world applications, and as such, require recognizing malware instances in an open-set scenario, i.e., some unknown families are also included in the test set, which has been rarely and nonthoroughly investigated in the cyber-security domain. One practical solution for MOSR may consider jointly classifying known and detecting unknown malware families by a single classifier (e.g., neural network) from the variance of the predicted probability distribution on known families. However, conventional well-trained classifiers usually tend to obtain overly high recognition probabilities in the outputs, especially when the instance feature distributions are similar to each other, e.g., unknown versus known malware families, and thus, dramatically degrade the recognition on novel unknown malware families. To address the problem and construct an applicable MOSR system, we propose a novel model that can conservatively synthesize malware instances to mimic unknown malware families and support a more robust training of the classifier. More specifically, we build upon the generative adversarial networks to explore and obtain marginal malware instances that are close to known families while falling into mimical unknown ones to guide the classifier to lower and flatten the recognition probabilities of unknown families and relatively raise that of known ones to rectify the performance of classification and detection. A cooperative training scheme involving the classification, synthesizing and rectification are further constructed to facilitate the training and jointly improve the model performance. Moreover, we also build a new large-scale malware dataset, named MAL-100, to fill the gap of lacking a large open-set malware benchmark dataset. Experimental results on two widely used malware datasets and our MAL-100 demonstrate the effectiveness of our model compared with other representative methods.

The prevalent personalized federated learning (PFL) usually pursues a trade-off between personalization and generalization by maintaining a shared global model to guide the training process of local models. However, the sole global model may easily transfer deviated context knowledge to some local models when multiple latent contexts exist across the local datasets. In this paper, we propose a novel concept called contextualized generalization (CG) to provide each client with fine-grained context knowledge that can better fit the local data distributions and facilitate faster model convergence, based on which we properly design a framework of PFL, dubbed CGPFL. We conduct detailed theoretical analysis, in which the convergence guarantee is presented and a speedup of order 1/2 w.r.t the number of contexts over most existing methods is granted. To quantitatively study the generalization-personalization trade-off, we introduce the generalization error measure and prove that the proposed CGPFL can achieve a better trade-off than existing solutions. Moreover, our theoretical analysis further inspires a heuristic algorithm to find a near-optimal trade-off in CGPFL. Experimental results on multiple real-world datasets show that our approach surpasses the state-of-the-art methods on test accuracy by a significant margin.

Modern machine learning (ML) applications are often deployed in the cloud environment to exploit the computational power of clusters. However, this in-cloud computing scheme cannot satisfy the demands of emerging edge intelligence scenarios, including providing personalized models, protecting user privacy, adapting to real-time tasks, and saving resource cost. In order to conquer the limitations of conventional in-cloud computing, there comes the rise of on-device learning, which makes the end-to-end ML procedure totally on user devices, without unnecessary involvement of the cloud. In spite of the promising advantages of on-device learning, implementing a high-performance on-device learning system still faces with many severe challenges, such as insufficient user training data, backward propagation (BP) blocking, and limited peak processing speed. Observing the substantial improvement space in the implementation and acceleration of on-device learning systems, we intend to present a comprehensive analysis of the latest research progress and point out potential optimization directions from the system perspective. This survey presents a software and hardware synergy of on-device learning techniques, covering the scope of model-level neural network design, algorithm-level training optimization, and hardware-level instruction acceleration. We hope this survey could bring fruitful discussions and inspire the researchers to further promote the field of edge intelligence.

Zero-shot learning aims at recognizing unseen classes (no training example) with knowledge transferred from seen classes. This is typically achieved by exploiting a semantic feature space shared by both seen and unseen classes, i.e., attribute or word vector, as the bridge. One common practice in zero-shot learning is to train a projection between the visual and semantic feature spaces with labeled seen classes examples. When inferring, this learned projection is applied to unseen classes and recognizes the class labels by some metrics. However, the visual and semantic feature spaces are mutually independent and have quite different manifold structures. Under such a paradigm, most existing methods easily suffer from the domain shift problem and weaken the performance of zero-shot recognition. To address this issue, we propose a novel model called AMS-SFE. It considers the alignment of manifold structures by semantic feature expansion. Specifically, we build upon an autoencoder-based model to expand the semantic features from the visual inputs. Additionally, the expansion is jointly guided by an embedded manifold extracted from the visual feature space of the data. Our model is the first attempt to align both feature spaces by expanding semantic features and derives two benefits: first, we expand some auxiliary features that enhance the semantic feature space; second and more importantly, we implicitly align the manifold structures between the visual and semantic feature spaces; thus, the projection can be better trained and mitigate the domain shift problem. Extensive experiments show significant performance improvement, which verifies the effectiveness of our model.

One non-negligible flaw of the convolutional neural networks (CNNs) based single image super-resolution (SISR) models is that most of them are not able to restore high-resolution (HR) images containing sufficient high-frequency information. Worse still, as the depth of CNNs increases, the training easily suffers from the vanishing gradients. These problems hinder the effectiveness of CNNs in SISR. In this paper, we propose the Dual-view Attention Networks to alleviate these problems for SISR. Specifically, we propose the local aware (LA) and global aware (GA) attentions to deal with LR features in unequal manners, which can highlight the high-frequency components and discriminate each feature from LR images in the local and global views, respectively. Furthermore, the local attentive residual-dense (LARD) block that combines the LA attention with multiple residual and dense connections is proposed to fit a deeper yet easy to train architecture. The experimental results verified the effectiveness of our model compared with other state-of-the-art methods.

In most recent years, zero-shot recognition (ZSR) has gained increasing attention in machine learning and image processing fields. It aims at recognizing unseen class instances with knowledge transferred from seen classes. This is typically achieved by exploiting a pre-defined semantic feature space (FS), i.e., semantic attributes or word vectors, as a bridge to transfer knowledge between seen and unseen classes. However, due to the absence of unseen classes during training, the conventional ZSR easily suffers from domain shift and hubness problems. In this paper, we propose a novel ZSR learning framework that can handle these two issues well by adaptively adjusting semantic FS. To the best of our knowledge, our work is the first to consider the adaptive adjustment of semantic FS in ZSR. Moreover, our solution can be formulated to a more efficient framework that significantly boosts the training. Extensive experiments show the remarkable performance improvement of our model compared with other existing methods.

Unsupervised learning is becoming more and more important recently. As one of its key components, the autoencoder (AE) aims to learn a latent feature representation of data which is more robust and discriminative. However, most AE based methods only focus on the reconstruction within the encoder-decoder phase, which ignores the inherent relation of data, i.e., statistical and geometrical dependence, and easily causes overfitting. In order to deal with this issue, we propose an Exclusivity Enhanced (EE) unsupervised feature learning approach to improve the conventional AE. To the best of our knowledge, our research is the first to utilize such exclusivity concept to cooperate with feature extraction within AE. Moreover, in this paper we also make some improvements to the stacked AE structure especially for the connection of different layers from decoders, this could be regarded as a weight initialization trial. The experimental results show that our proposed approach can achieve remarkable performance compared with other related methods.

Zero-shot learning (ZSL) aims at recognizing unseen classes with knowledge transferred from seen classes. This is typically achieved by exploiting a semantic feature space (FS) shared by both seen and unseen classes, i.e., attributes or word vectors, as the bridge. However, due to the mutually disjoint of training (seen) and testing (unseen) data, existing ZSL methods easily and commonly suffer from the domain shift problem. To address this issue, we propose a novel model called AMS-SFE. It considers the Alignment of Manifold Structures by Semantic Feature Expansion. Specifically, we build up an autoencoder based model to expand the semantic features and joint with an alignment to an embedded manifold extracted from the visual FS of data. It is the first attempt to align these two FSs by way of expanding semantic features. Extensive experiments show the remarkable performance improvement of our model compared with other existing methods.

The Support Vector Machine(SVM) is well known in machine learning and artificial intelligence for its high performance in data classification, regression and forecasting. Usually for large scaled dataset, an incremental training algorithm is applied for tuning or balancing the training cost and the accuracy in SVM applications. This paper presents an improved incremental training approach for large scaled dataset on SVM. We focus on data's own distribution information to unfold our research, we proposed a self adaptive clustering method to extract the area and density information of data, a border detection technologies and uncertainty strategy is applied to maintain the border and some potential samples. Our proposed method can greatly reduce the training error for incremental training on SVM, especially for some uneven distribution dataset. We can greatly tuning or balancing the training cost and the accuracy of algorithms to achieve a better performance.