Room layout estimation from multiple-perspective images is poorly investigated due to the complexities that emerge from multi-view geometry, which requires muti-step solutions such as camera intrinsic and extrinsic estimation, image matching, and triangulation. However, in 3D reconstruction, the advancement of recent 3D foundation models such as DUSt3R has shifted the paradigm from the traditional multi-step structure-from-motion process to an end-to-end single-step approach. To this end, we introduce Plane-DUSt3R, a novel method for multi-view room layout estimation leveraging the 3D foundation model DUSt3R. Plane-DUSt3R incorporates the DUSt3R framework and fine-tunes on a room layout dataset (Structure3D) with a modified objective to estimate structural planes. By generating uniform and parsimonious results, Plane-DUSt3R enables room layout estimation with only a single post-processing step and 2D detection results. Unlike previous methods that rely on single-perspective or panorama image, Plane-DUSt3R extends the setting to handle multiple-perspective images. Moreover, it offers a streamlined, end-to-end solution that simplifies the process and reduces error accumulation. Experimental results demonstrate that Plane-DUSt3R not only outperforms state-of-the-art methods on the synthetic dataset but also proves robust and effective on in the wild data with different image styles such as cartoon. Our code is available at: https://github.com/justacar/Plane-DUSt3R

Scaling Transformer to a large scale without using some technical tricks such as learning rate warump and an obviously lower learning rate, is an extremely challenging task, and is increasingly gaining more attention. In this paper, we provide a theoretical analysis for the process of training Transformer and reveal a key problem behind model crash phenomenon in the training process, termed *spectral energy concentration* of ${W_q}^{\top} W_k$, which is the reason for a malignant entropy collapse, where ${W_q}$ and $W_k$ are the projection matrices for the query and the key in Transformer, respectively. To remedy this problem, motivated by *Weyl's Inequality*, we present a novel optimization strategy, \ie, making the weight updating in successive steps steady---if the ratio $\frac{\sigma_{1}(\nabla W_t)}{\sigma_{1}(W_{t-1})}$ is larger than a threshold, we will automatically bound the learning rate to a weighted multiple of $\frac{\sigma_{1}(W_{t-1})}{\sigma_{1}(\nabla W_t)}$, where $\nabla W_t$ is the updating quantity in step $t$. Such an optimization strategy can prevent spectral energy concentration to only a few directions, and thus can avoid malignant entropy collapse which will trigger the model crash. We conduct extensive experiments using ViT, Swin-Transformer and GPT, showing that our optimization strategy can effectively and stably train these (Transformer) models without using learning rate warmup.

Recent advances in training-free attention control methods have enabled flexible and efficient text-guided editing capabilities for existing image and video generation models. However, current approaches struggle to simultaneously deliver strong editing strength while preserving consistency with the source. For instance, in color-editing tasks, they struggle to maintain structural consistency in edited regions while preserving the rest intact. This limitation becomes particularly critical in multi-round and video editing, where visual errors can accumulate over time. Moreover, most existing methods enforce global consistency, which limits their ability to modify individual attributes such as texture while preserving others, thereby hindering fine-grained editing. Recently, the architectural shift from U-Net to Multi-Modal Diffusion Transformers (MM-DiT) has brought significant improvements in generative performance and introduced a novel mechanism for integrating text and vision modalities. These advancements pave the way for overcoming challenges that previous methods failed to resolve. Through an in-depth analysis of MM-DiT, we identify three key insights into its attention mechanisms. Building on these, we propose ConsistEdit, a novel attention control method specifically tailored for MM-DiT. ConsistEdit incorporates vision-only attention control, mask-guided pre-attention fusion, and differentiated manipulation of the query, key, and value tokens to produce consistent, prompt-aligned edits. Extensive experiments demonstrate that ConsistEdit achieves state-of-the-art performance across a wide range of image and video editing tasks, including both structure-consistent and structure-inconsistent scenarios. Unlike prior methods, it is the first approach to perform editing across all inference steps and attention layers without handcraft, significantly enhancing reliability and consistency, which enables robust multi-round and multi-region editing. Furthermore, it supports progressive adjustment of structural consistency, enabling finer control. ConsistEdit represents a significant advancement in generative model editing and unlocks the full editing potential of MM-DiT architectures.

The advent of large pre-trained models has brought about a paradigm shift in both visual representation learning and natural language processing. However, clustering unlabeled images, as a fundamental and classic machine learning problem, still lacks an effective solution, particularly for large-scale datasets. In this paper, we propose a novel image clustering pipeline that leverages the powerful feature representation of large pre-trained models such as CLIP and cluster images effectively and efficiently at scale. We first developed a novel algorithm to estimate the number of clusters in a given dataset. We then show that the pre-trained features are significantly more structured by further optimizing the rate reduction objective. The resulting features may significantly improve the clustering accuracy, e.g., from 57\% to 66\% on ImageNet-1k. Furthermore, by leveraging CLIP's multimodality bridge between image and text, we develop a simple yet effective self-labeling algorithm that produces meaningful text labels for the clusters. Through extensive experiments, we show that our pipeline works well on standard datasets such as CIFAR-10, CIFAR-100, and ImageNet-1k. It also extends to datasets without predefined labels, such as LAION-Aesthetics and WikiArts.

Deep metric learning (DML) aims to learn a discriminative high-dimensional embedding space for downstream tasks like classification, clustering, and retrieval. Prior literature predominantly focuses on pair-based and proxy-based methods to maximize inter-class discrepancy and minimize intra-class diversity. However, these methods tend to suffer from the collapse of the embedding space due to their over-reliance on label information. This leads to sub-optimal feature representation and inferior model performance. To maintain the structure of embedding space and avoid feature collapse, we propose a novel loss function called Anti-Collapse Loss. Specifically, our proposed loss primarily draws inspiration from the principle of Maximal Coding Rate Reduction. It promotes the sparseness of feature clusters in the embedding space to prevent collapse by maximizing the average coding rate of sample features or class proxies. Moreover, we integrate our proposed loss with pair-based and proxy-based methods, resulting in notable performance improvement. Comprehensive experiments on benchmark datasets demonstrate that our proposed method outperforms existing state-of-the-art methods. Extensive ablation studies verify the effectiveness of our method in preventing embedding space collapse and promoting generalization performance.

Recently, deep metric learning (DML) has achieved great success. Some existing DML methods propose adaptive sample mining strategies, which learn to weight the samples, leading to interesting performance. However, these methods suffer from a small memory (e.g., one training batch), limiting their efficacy. In this work, we introduce a data-driven method, meta-mining strategy with semiglobal information (MMSI), to apply meta-learning to learn to weight samples during the whole training, leading to an adaptive mining strategy. To introduce richer information than one training batch only, we elaborately take advantage of the validation set of meta-learning by implicitly adding additional validation sample information to training. Furthermore, motivated by the latest self-supervised learning, we introduce a dictionary (memory) that maintains very large and diverse information. Together with the validation set, this dictionary presents much richer information to the training, leading to promising performance. In addition, we propose a new theoretical framework that can formulate pairwise and tripletwise metric learning loss functions in a unified framework. This framework brings new insights to society and facilitates us to generalize our MMSI to many existing DML methods. We conduct extensive experiments on three public datasets, CUB200-2011, Cars-196, and Stanford Online Products (SOP). Results show that our method can achieve the state of the art or very competitive performance. Our source codes have been made available at https://github.com/NUST-Machine-Intelligence-Laboratory/MMSI.

We consider the problem of simultaneously clustering and learning a linear representation of data lying close to a union of low-dimensional manifolds, a fundamental task in machine learning and computer vision. When the manifolds are assumed to be linear subspaces, this reduces to the classical problem of subspace clustering, which has been studied extensively over the past two decades. Unfortunately, many real-world datasets such as natural images can not be well approximated by linear subspaces. On the other hand, numerous works have attempted to learn an appropriate transformation of the data, such that data is mapped from a union of general non-linear manifolds to a union of linear subspaces (with points from the same manifold being mapped to the same subspace). However, many existing works have limitations such as assuming knowledge of the membership of samples to clusters, requiring high sampling density, or being shown theoretically to learn trivial representations. In this paper, we propose to optimize the Maximal Coding Rate Reduction metric with respect to both the data representation and a novel doubly stochastic cluster membership, inspired by state-of-the-art subspace clustering results. We give a parameterization of such a representation and membership, allowing efficient mini-batching and one-shot initialization. Experiments on CIFAR-10, -20, -100, and TinyImageNet-200 datasets show that the proposed method is much more accurate and scalable than state-of-the-art deep clustering methods, and further learns a latent linear representation of the data.4

Despite strong empirical performance for image classification, deep neural networks are often regarded as ``black boxes'' and they are difficult to interpret. On the other hand, sparse convolutional models, which assume that a signal can be expressed by a linear combination of a few elements from a convolutional dictionary, are powerful tools for analyzing natural images with good theoretical interpretability and biological plausibility. However, such principled models have not demonstrated competitive performance when compared with empirically designed deep networks. This paper revisits the sparse convolutional modeling for image classification and bridges the gap between good empirical performance (of deep learning) and good interpretability (of sparse convolutional models). Our method uses differentiable optimization layers that are defined from convolutional sparse coding as drop-in replacements of standard convolutional layers in conventional deep neural networks. We show that such models have equally strong empirical performance on CIFAR-10, CIFAR-100 and ImageNet datasets when compared to conventional neural networks. By leveraging stable recovery property of sparse modeling, we further show that such models can be much more robust to input corruptions as well as adversarial perturbations in testing through a simple proper trade-off between sparse regularization and data reconstruction terms.

Single-image room layout reconstruction aims to reconstruct the enclosed 3D structure of a room from a single image. Most previous work relies on the cuboid shape prior. This paper considers a more general indoor assumption, i.e., the room layout consists of a single ceiling, a single floor, and several vertical walls. To this end, we first employ Convolutional Neural Networks to detect planes and vertical lines between adjacent walls. Meanwhile, estimating the 3D parameters for each plane. Then, a simple yet effective geometric reasoning method is adopted to achieve room layout reconstruction. Furthermore, we optimize the 3D plane parameters to reconstruct a geometrically consistent room layout between planes and lines. The experimental results on public datasets validate the effectiveness and efficiency of our method.

With the breakthrough of deep learning, lip reading technologies are under extraordinarily rapid progress. It is well-known that Chinese is the most widely spoken language in the world. Unlike alphabetic languages, it involves more than 1,000 pronunciations as Pinyin, and nearly 90,000 pictographic characters as Hanzi, which makes lip reading of Chinese very challenging. In this paper, we implement visual-only Chinese lip reading of unconstrained sentences in a two-step end-to-end architecture (LipCH-Net), in which two deep neural network models are employed to perform the recognition of Pictureto-Pinyin (mouth motion pictures to pronunciations) and the recognition of Pinyin-to-Hanzi (pronunciations to texts) respectively, before having a jointly optimization to improve the overall performance. In addition, two modules in the Pinyin-to-Hanzi model are pre-trained separately with large auxiliary data in advance of sequence-to-sequence training to make the best of long sequence matches for avoiding ambiguity. We collect 6-month daily news broadcasts from China Central Television (CCTV) website, and semi-automatically label them into a 20.95 GB dataset with 20,495 natural Chinese sentences. When trained on the CCTV dataset, the LipCH-Net model outperforms the performance of all stateof-the-art lip reading frameworks. According to the results, our scheme not only accelerates training and reduces overfitting, but also overcomes syntactic ambiguity of Chinese which provides a baseline for future relevant work.

Cloud computing opens a new door for designing the next-generation video distribution platform. As video services move to the cloud, some related data-intensive applications, such as recommender system, automatic scoring mechanism, and prediction algorithm, should also be cloud-based. Since traditional cloud file systems like MapReduce/Hadoop exhibit cost disadvantages in data accessing, a Cache A Replica On Modification (CAROM) cloud file system is designed to achieve high data availability and low storage cost, which provides resiliency in cloud file systems with high efficiency. In this paper, we focus on moving data-intensive applications to the CAROM cloud file system, for optimizing access latencies while maintaining the benefit of low storage cost. To achieve this, we propose a novel scheduling mechanism as a lubricant between CAROM and data-intensive applications. Our scheme consists of three parts. First, tripartite graph is employed to describe the relationships among tasks, computation nodes, and data nodes. Second, we give a 1:1:1 framework based on the situation that the data of task have been stored in the cache, and introduce two variant frameworks 1:1:M and 1:N:M with the consideration of limitations of cache size and performance of task. Finally, a k-list algorithm is proposed as an approximation algorithm, and its mathematical definitions and proofs are given in detail. We conduct simulations to evaluate our scheme and the results prove that the performance of our algorithm is significantly better than that of the general two-layer algorithm.

To facilitate the safe and comfortable driving, vehicular ad hoc networks (VANETs) will be flooded with plenty of multimedia files, such as images, music and video clips. However, due to the dynamic and transient contacts between moving vehicles, these multimedia files distribution over VANETs often involves transmission failure and terrible user experience. In this paper, we propose an efficient infrastructure-less multimedia publish/subscribe scheme for an urban area. In cities, there are lots of parked vehicles, presenting as parking clusters, owning the ability of calculation, storage and communication. Our scheme relies on these parking clusters to cache and distribute the multimedia files for moving users. For each subscription, the parking cluster distributes the file chunks to the subscriber during their contact time. For the remained content chunks, the parking cluster will distribute them to slave vehicles that have no downloading request. Then, the slave vehicles transfer the received file chunks to a parking cluster, where the subscriber can continue the unfinished downloading when it drives through. Theoretical results illustrate the effectiveness of our approach and extensive simulations results demonstrate that the proposed scheme achieves a higher downloading ratio with different file sizes, especially in sparse traffic and multiple subscribers conditions.