Kun Zhou

Wenbo Li, Zhe Lin, Kun Zhou et al.

Recent studies have shown the importance of modeling long-range interactions in the inpainting problem. To achieve this goal, existing approaches exploit either standalone attention techniques or transformers, but usually under a low resolution in consideration of computational cost. In this paper, we present a novel transformer-based model for large hole inpainting, which unifies the merits of transformers and convolutions to efficiently process high-resolution images. We carefully design each component of our framework to guarantee the high fidelity and diversity of recovered images. Specifically, we customize an inpainting-oriented transformer block, where the attention module aggregates non-local information only from partial valid tokens, indicated by a dynamic mask. Extensive experiments demonstrate the state-of-the-art performance of the new model on multiple benchmark datasets. Code is released at https://github.com/fenglinglwb/MAT.

Wenbo Li, Xin Yu, Kun Zhou et al.

Generative adversarial networks (GANs) have made great success in image inpainting yet still have difficulties tackling large missing regions. In contrast, iterative probabilistic algorithms, such as autoregressive and denoising diffusion models, have to be deployed with massive computing resources for decent effect. To achieve high-quality results with low computational cost, we present a novel pixel spread model (PSM) that iteratively employs decoupled probabilistic modeling, combining the optimization efficiency of GANs with the prediction tractability of probabilistic models. As a result, our model selectively spreads informative pixels throughout the image in a few iterations, largely enhancing the completion quality and efficiency. On multiple benchmarks, we achieve new state-of-the-art performance. Code is released at https://github.com/fenglinglwb/PSM.

Xiaolei Wang, Kun Zhou, Ji-Rong Wen et al.

Conversational recommender systems (CRS) aim to proactively elicit user preference and recommend high-quality items through natural language conversations. Typically, a CRS consists of a recommendation module to predict preferred items for users and a conversation module to generate appropriate responses. To develop an effective CRS, it is essential to seamlessly integrate the two modules. Existing works either design semantic alignment strategies, or share knowledge resources and representations between the two modules. However, these approaches still rely on different architectures or techniques to develop the two modules, making it difficult for effective module integration. To address this problem, we propose a unified CRS model named UniCRS based on knowledge-enhanced prompt learning. Our approach unifies the recommendation and conversation subtasks into the prompt learning paradigm, and utilizes knowledge-enhanced prompts based on a fixed pre-trained language model (PLM) to fulfill both subtasks in a unified approach. In the prompt design, we include fused knowledge representations, task-specific soft tokens, and the dialogue context, which can provide sufficient contextual information to adapt the PLM for the CRS task. Besides, for the recommendation subtask, we also incorporate the generated response template as an important part of the prompt, to enhance the information interaction between the two subtasks. Extensive experiments on two public CRS datasets have demonstrated the effectiveness of our approach.

Kaizhang Kang, Chong Zeng, Hongzhi Wu et al. · stanford

We present a novel framework to automatically learn to transform the differential cues from a stack of images densely captured with a rotational motion into spatially discriminative and view-invariant per-pixel features at each view. These low-level features can be directly fed to any existing multi-view stereo technique for enhanced 3D reconstruction. The lighting condition during acquisition can also be jointly optimized in a differentiable fashion. We sample from a dozen of pre-scanned objects with a wide variety of geometry and reflectance to synthesize a large amount of high-quality training data. The effectiveness of our features is demonstrated on a number of challenging objects acquired with a lightstage, comparing favorably with state-of-the-art techniques. Finally, we explore additional applications of geometric detail visualization and computational stylization of complex appearance.

Kun Zhou, Wenbo Li, Xiaoguang Han et al.

For video frame interpolation (VFI), existing deep-learning-based approaches strongly rely on the ground-truth (GT) intermediate frames, which sometimes ignore the non-unique nature of motion judging from the given adjacent frames. As a result, these methods tend to produce averaged solutions that are not clear enough. To alleviate this issue, we propose to relax the requirement of reconstructing an intermediate frame as close to the GT as possible. Towards this end, we develop a texture consistency loss (TCL) upon the assumption that the interpolated content should maintain similar structures with their counterparts in the given frames. Predictions satisfying this constraint are encouraged, though they may differ from the pre-defined GT. Without the bells and whistles, our plug-and-play TCL is capable of improving the performance of existing VFI frameworks. On the other hand, previous methods usually adopt the cost volume or correlation map to achieve more accurate image/feature warping. However, the O(N^2) ({N refers to the pixel count}) computational complexity makes it infeasible for high-resolution cases. In this work, we design a simple, efficient (O(N)) yet powerful cross-scale pyramid alignment (CSPA) module, where multi-scale information is highly exploited. Extensive experiments justify the efficiency and effectiveness of the proposed strategy.

Kun Zhou, KenKun Liu, Wenbo Li et al.

Previous studies show the necessity of global and local adjustment for image enhancement. However, existing convolutional neural networks (CNNs) and transformer-based models face great challenges in balancing the computational efficiency and effectiveness of global-local information usage. Especially, existing methods typically adopt the global-to-local fusion mode, ignoring the importance of bidirectional interactions. To address those issues, we propose a novel mutual guidance network (MGN) to perform effective bidirectional global-local information exchange while keeping a compact architecture. In our design, we adopt a two-branch framework where one branch focuses more on modeling global relations while the other is committed to processing local information. Then, we develop an efficient attention-based mutual guidance approach throughout our framework for bidirectional global-local interactions. As a result, both the global and local branches can enjoy the merits of mutual information aggregation. Besides, to further refine the results produced by our MGN, we propose a novel residual integration scheme following the divide-and-conquer philosophy. The extensive experiments demonstrate the effectiveness of our proposed method, which achieves state-of-the-art performance on several public image enhancement benchmarks.

Yifan Du, Yuqi Huo, Kun Zhou et al.

Video Multimodal Large Language Models (MLLMs) have shown remarkable capability of understanding the video semantics on various downstream tasks. Despite the advancements, there is still a lack of systematic research on visual context representation, which refers to the scheme to select frames from a video and further select the tokens from a frame. In this paper, we explore the design space for visual context representation, and aim to improve the performance of video MLLMs by finding more effective representation schemes. Firstly, we formulate the task of visual context representation as a constrained optimization problem, and model the language modeling loss as a function of the number of frames and the number of embeddings (or tokens) per frame, given the maximum visual context window size. Then, we explore the scaling effects in frame selection and token selection respectively, and fit the corresponding function curve by conducting extensive empirical experiments. We examine the effectiveness of typical selection strategies and present empirical findings to determine the two factors. Furthermore, we study the joint effect of frame selection and token selection, and derive the optimal formula for determining the two factors. We demonstrate that the derived optimal settings show alignment with the best-performed results of empirical experiments. Our code and model are available at: https://github.com/RUCAIBox/Opt-Visor.

Jingming Liu, Yumeng Li, Boyuan Xiao et al.

Under pure textual modality, Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning tasks by decomposing them into simpler sub-problems. However, Multimodal Large Language Models (MLLMs) still struggle with some seemingly straightforward visual tasks, such as counting and solving jigsaw puzzles. We argue that these tasks challenge the ability of visual-to-textual conversion, where MLLMs convert visual information perceived from the input scene, to textual information for further reasoning and generating the answer. If the complexity of the visual input is beyond the perceptual capability of the MLLMs, without decomposing this conversion process, simply scaling inference-time reasoning cannot solve the task because it repeatedly encounters the same perceptual bottleneck. We propose an approach, autonomous imagination, to enable MLLMs to iteratively modify visual inputs (e.g. isolating objects, rearranging puzzle pieces) into intermediate visual states, decomposing visual-to-textual conversion into closed-loop visual modification steps. We show that, without any retraining, MLLMs can now solve tasks initially beyond their perceptual capability, highlighting that closed-loop visual modification can be an effective way of decomposing the visual reasoning task into solvable substeps. Our code and data are released at https://future-item.github.io/autoimagine-site/.

Kun Zhou, Wenbo Li, Liying Lu et al.

Long-range temporal alignment is critical yet challenging for video restoration tasks. Recently, some works attempt to divide the long-range alignment into several sub-alignments and handle them progressively. Although this operation is helpful in modeling distant correspondences, error accumulation is inevitable due to the propagation mechanism. In this work, we present a novel, generic iterative alignment module which employs a gradual refinement scheme for sub-alignments, yielding more accurate motion compensation. To further enhance the alignment accuracy and temporal consistency, we develop a non-parametric re-weighting method, where the importance of each neighboring frame is adaptively evaluated in a spatial-wise way for aggregation. By virtue of the proposed strategies, our model achieves state-of-the-art performance on multiple benchmarks across a range of video restoration tasks including video super-resolution, denoising and deblurring. Our project is available in \url{https://github.com/redrock303/Revisiting-Temporal-Alignment-for-Video-Restoration.git}.

Wenbo Li, Kun Zhou, Lu Qi et al.

Single image super-resolution (SISR) deals with a fundamental problem of upsampling a low-resolution (LR) image to its high-resolution (HR) version. Last few years have witnessed impressive progress propelled by deep learning methods. However, one critical challenge faced by existing methods is to strike a sweet spot of deep model complexity and resulting SISR quality. This paper addresses this pain point by proposing a linearly-assembled pixel-adaptive regression network (LAPAR), which casts the direct LR to HR mapping learning into a linear coefficient regression task over a dictionary of multiple predefined filter bases. Such a parametric representation renders our model highly lightweight and easy to optimize while achieving state-of-the-art results on SISR benchmarks. Moreover, based on the same idea, LAPAR is extended to tackle other restoration tasks, e.g., image denoising and JPEG image deblocking, and again, yields strong performance. The code is available at https://github.com/dvlab-research/Simple-SR.

Hongfei Zhang, Kun Zhou, Ruizheng Wu et al.

Image dehazing remains a challenging problem due to the spatially varying nature of haze in real-world scenes. While existing methods have demonstrated the promise of large-scale pretrained models for image dehazing, their architecture-specific designs hinder adaptability across diverse scenarios with different accuracy and efficiency requirements. In this work, we systematically investigate the generalization capability of pretrained depth representations-learned from millions of diverse images-for image dehazing. Our empirical analysis reveals that the learned deep depth features maintain remarkable consistency across varying haze levels. Building on this insight, we propose a plug-and-play RGB-D fusion module that seamlessly integrates with diverse dehazing architectures. Extensive experiments across multiple benchmarks validate both the effectiveness and broad applicability of our approach.

Jiawei Lu, He Wang, Tianjia Shao et al.

Generating new images with desired properties (e.g. new view/poses) from source images has been enthusiastically pursued recently, due to its wide range of potential applications. One way to ensure high-quality generation is to use multiple sources with complementary information such as different views of the same object. However, as source images are often misaligned due to the large disparities among the camera settings, strong assumptions have been made in the past with respect to the camera(s) or/and the object in interest, limiting the application of such techniques. Therefore, we propose a new general approach which models multiple types of variations among sources, such as view angles, poses, facial expressions, in a unified framework, so that it can be employed on datasets of vastly different nature. We verify our approach on a variety of data including humans bodies, faces, city scenes and 3D objects. Both the qualitative and quantitative results demonstrate the better performance of our method than the state of the art.

Wenbo Li, Kun Zhou, Lu Qi et al.

We consider the single image super-resolution (SISR) problem, where a high-resolution (HR) image is generated based on a low-resolution (LR) input. Recently, generative adversarial networks (GANs) become popular to hallucinate details. Most methods along this line rely on a predefined single-LR-single-HR mapping, which is not flexible enough for the SISR task. Also, GAN-generated fake details may often undermine the realism of the whole image. We address these issues by proposing best-buddy GANs (Beby-GAN) for rich-detail SISR. Relaxing the immutable one-to-one constraint, we allow the estimated patches to dynamically seek the best supervision during training, which is beneficial to producing more reasonable details. Besides, we propose a region-aware adversarial learning strategy that directs our model to focus on generating details for textured areas adaptively. Extensive experiments justify the effectiveness of our method. An ultra-high-resolution 4K dataset is also constructed to facilitate future super-resolution research.

Kaizhang Kang, Cihui Xie, Ruisheng Zhu et al.

We present a novel framework to learn to convert the perpixel photometric information at each view into spatially distinctive and view-invariant low-level features, which can be plugged into existing multi-view stereo pipeline for enhanced 3D reconstruction. Both the illumination conditions during acquisition and the subsequent per-pixel feature transform can be jointly optimized in a differentiable fashion. Our framework automatically adapts to and makes efficient use of the geometric information available in different forms of input data. High-quality 3D reconstructions of a variety of challenging objects are demonstrated on the data captured with an illumination multiplexing device, as well as a point light. Our results compare favorably with state-of-the-art techniques.

Junyan Wang, Sai-kit Yeung, Jue Wang et al.

Recent works on interactive video object cutout mainly focus on designing dynamic foreground-background (FB) classifiers for segmentation propagation. However, the research on optimally removing errors from the FB classification is sparse, and the errors often accumulate rapidly, causing significant errors in the propagated frames. In this work, we take the initial steps to addressing this problem, and we call this new task \emph{segmentation rectification}. Our key observation is that the possibly asymmetrically distributed false positive and false negative errors were handled equally in the conventional methods. We, alternatively, propose to optimally remove these two types of errors. To this effect, we propose a novel bilayer Markov Random Field (MRF) model for this new task. We also adopt the well-established structured learning framework to learn the optimal model from data. Additionally, we propose a novel one-class structured SVM (OSSVM) which greatly speeds up the structured learning process. Our method naturally extends to RGB-D videos as well. Comprehensive experiments on both RGB and RGB-D data demonstrate that our simple and effective method significantly outperforms the segmentation propagation methods adopted in the state-of-the-art video cutout systems, and the results also suggest the potential usefulness of our method in image cutout system.