Stereo matching recovers 3D scene information based on the correlation between corresponding pixels. Despite impressive progress, existing methods lack sufficient correlation priors in ill-posed regions such as occlusions, detailed and reflective regions. In this paper, we propose Geometry Aware Stereo Matching Network (GEAStereo) to enhance geometric structure perception and address this issue. We adaptively incorporate the Monocular Disparity Distribution Prior into the stereo cost volume, building Mono-Stereo Fusion Volume (MSFV), which effectively captures global geometric structures and rectifies the correlation information in ill-posed regions. Furthermore, we introduce rich detail information from gradient features and construct a Detail-Aware Volume (DAV) by aggregating the group-wise cost volume under the guidance of gradient spatial attention, thus enhancing the correlation modeling in detailed structures. Jointly, MSFV and DAV provide rich correlation priors for disparity iterative optimization. Experimental results show that our method achieves competitive results on the ETH3D and KITTI2015 benchmarks. Compared with the state-of-the-art methods, our method demonstrates stronger performance in zero-shot generalization.

Despite significant advancements in diffusion models for generating high-quality portrait images, the problem of malformed hands remains a critical and unresolved challenge. Existing methods primarily focus on structural restoration, often neglecting the semantic coherence between the corrected hand region and the overall image. To address this limitation, we propose a novel dual-pathway diffusion malformed hand correction method, named DD-MHC, which integrates a Global Context Aware Module (GCAM) and a Local Structure Refinement Module (LSRM) through a dual-path architecture. Under the synergy of cross-attention and spatial attention, these modules can effectively fuse global contextual features and local guiding cues, enabling precise restoration of the hand region. Comprehensive experiments demonstrate that DD-MHC significantly outperforms existing competing methods, particularly in enhancing the semantic consistency between the corrected hand region and the overall image. Additionally, to address the challenge of data scarcity in the research on correcting malformed hands within unconstrained scene portraits, we construct a brand-new General-Scene Portrait Dataset (GSPD), providing a standardized and reproducible data platform for subsequent related research.

Depth estimation plays a crucial role in 3D scene understanding and is extensively used in a wide range of vision tasks. Image-based methods struggle in challenging scenarios, while event cameras offer high dynamic range and temporal resolution but face difficulties with sparse data. Combining event and image data provides significant advantages, yet effective integration remains challenging. Existing CNN-based fusion methods struggle with occlusions and depth disparities due to limited receptive fields, while Transformer-based fusion methods often lack deep modality interaction. To address these issues, we propose UniCT Depth, an event-image fusion method that unifies CNNs and Transformers to model local and global features. We propose the Convolution-compensated ViT Dual SA (CcViT-DA) Block, designed for the encoder, which integrates Context Modeling Self-Attention (CMSA) to capture spatial dependencies and Modal Fusion Self-Attention (MFSA) for effective cross-modal fusion. Furthermore, we design the tailored Detail Compensation Convolution (DCC) Block to improve texture details and enhances edge representations. Extensive experiments show that UniCT Depth outperforms existing image, event, and fusion-based monocular depth estimation methods across key metrics.

Unsupervised domain adaptation (UDA) has been widely studied with the goal of transferring knowledge from a label-rich source domain to a related but unlabeled target domain. Most UDA techniques achieve this by reducing the feature discrepancies between the two domains to learn domain-invariant feature representations. While domain-invariant feature representations can reduce the differences between the source and target domains, excessively simplifying these differences may cause the model to overlook important domain-specific features, resulting in a decline in transfer learning effectiveness. To address this issue, this paper proposes a novel Gaussian Mixture Model for graph domain adaptation (GMM). This model effectively reduces the distributional bias between the source and target domains by modeling the distribution differences on a graph structure. GMM leverages the local structural information of the graph and the clustering capability of the Gaussian mixture model to automatically learn the latent mapping relationships between the source and target domains. To the best of our knowledge, this is the first work to introduce a Gaussian mixture model into UDA. Extensive experimental results on three standard benchmarks demonstrate that the proposed GMM algorithm outperforms state-of-the-art unsupervised domain adaptation methods in terms of performance.

Event cameras have the potential to overcome the limitations of classical computer vision in real-world applications. Depth estimation is a crucial step for high-level robotics tasks and has attracted much attention from the community. In this paper, we propose an event-based dense depth estimation architecture, Mixed-EF2DNet, which firstly predicts inter-grid optical flow to compensate for lost temporal information, and then estimates multiple contextual depth maps that are fused to generate a robust depth estimation map. To supervise the network training, we further design a smoothing loss function used to smooth local depth estimates and facilitate estimating reasonable depth for pixels without events. In addition, we introduce SE-resblocks in the depth network to enhance the network representation by selecting feature channels. Experimental evaluations on both real-world and synthetic datasets show that our method performs better in terms of accuracy when compared to state-of-the-art algorithms, especially in scene detail estimation. Besides, our method demonstrates excellent generalization in cross-dataset tasks.

As a long-standing task, image restoration aims to recover the latent sharp image from its degraded counterpart. In recent years, owing to the strong ability of self-attention in capturing long-range dependencies, Transformer based methods have achieved promising performance on multifarious image restoration tasks. However, the canonical self-attention leads to quadratic complexity with respect to input size, hindering its further applications in image restoration. In this paper, we propose a Strip Attention Network (SANet) for image restoration to integrate information in a more efficient and effective manner. Specifically, a strip attention unit is proposed to harvest the contextual information for each pixel from its adjacent pixels in the same row or column. By employing this operation in different directions, each location can perceive information from an expanded region. Furthermore, we apply various receptive fields in different feature groups to enhance representation learning. Incorporating these designs into a U-shaped backbone, our SANet performs favorably against state-of-the-art algorithms on several image restoration tasks. The code is available at https://github.com/c-yn/SANet.

The recently emerging event camera has grown into a new type of sensor in the realm of vision, with benefits such as low power consumption, high dynamic range (HDR), microsecond time resolution, and no motion blur. While event cameras offer numerous advantages over conventional cameras, they only capture changes in intensity and give up lots of environmental details. This paper proposes an end-to-end UNet network called SCSE-E2VID to synthesize gray images from asynchronous events. We design an event fusion block to feed more related events to the encoder, allowing the network to extract more valuable features. The famous attention module called Spatial and Channel ‘Squeeze & Excitation’ Block (SCSE) is utilized to remove artifacts and better extract spatiotemporal features for the decoder. Besides, we add parallel convolutions in the upsampling block and refine the output features, which supplement content in reduced channels. In order to evaluate the performance of our proposed SCSE-E2VID, we implement quantitative and qualitative comparisons based on the public IJRR and HQF datasets. The results show that our method achieves better performance in terms of perceptual similarity and structural similarity when compared with state-of-art methods and demonstrates comparable performance in terms of squared error.

Optical flow estimation is an essential task in computer vision. Standard cameras are prone to blurred images or over-saturated regions under extreme conditions. The event camera is a novel vision sensor inspired by the biological retina. It has the advantages of high time resolution, low delay and high dynamic range. We propose a new event representation and a novel deep learning network E-HANet (Event-based Hybrid Attention Network) for event-based dense optical flow estimation. To take full advantage of the complementarity between positive and negative events, we introduce the stacked positive and negative event slices as input. The feature extractor based on the channel attention is able to model the features from different event slices and fuse them by weight. We then present the hybrid attention weighting module to globally aggregate motion features. Compared to the event-based state-of-the-art, our approach reduces the average end-point error by 3% on the MVSEC dataset and 20% on the DSEC-Flow dataset.

Event cameras are bio-inspired sensors that respond to pixel-level brightness changes in the form of asynchronous and sparse events. Each event is a 4-D tuple, i.e., (timestamp, x, y, polarity). Recently, algorithms for object detection based on learning have made considerable strides for event cameras. These methods transform event sequences into a tensor-like representation which can be processed by deep learning methods. However, these conversion methods don’t make full use of the polarity information of event sequences. We discover that tensor-like representations of different polarities show different features. Thus, we propose Polar Loss, a loss function that can enhance the difference between tensor-like representations of different polarities. To evaluate the effectiveness of our loss function, we design a network architecture for object detection. Our results state that our model shows good flexibility and expressiveness on event-based object detection when trained with the Polar Loss.

Recently, the emerging event camera shows great potential for robotics and AR/VR application thanks to its advantages including low latency, high dynamic range, low power consumption, etc. However, the output of event cameras usually has a large amount of noise which will affect unlocking their potentials and obtaining wider applications. In this paper, we present a two-stage local spatio-temporal event filter (LSTEF) based on adaptive thresholds. Considering the spatio-temporal constraints of generated event streams, a local sliding window is adopted for the noise candidate selection stage and noise filtering stage. An adaptive thresholding mechanism is also introduced into the filter in order to improve the generalization performance. Corresponding experimental evaluations are performed on the public datasets to prove the efficiency of the proposed filter. Results show that the presented LSTEF can successfully achieve the event denoising, and at the same time, effectively preserve useful scene information.