Wide-angle cameras, despite their popularity for content creation, suffer from distortion-induced facial stretching—especially at the edge of the lens—which degrades visual appeal. To address this issue, we propose a structure-to-detail portrait correction model named ImagePC. It integrates the long-range awareness of the transformer and multi-step denoising of diffusion models into a unified framework, achieving global structural robustness and local detail refinement. Besides, considering the high cost of obtaining video labels, we then repurpose ImagePC for unlabeled wide-angle videos (termed VideoPC), by spatiotemporal diffusion adaption with spatial consistency and temporal smoothness constraints. For the former, we encourage the denoised image to approximate pseudo labels following the wide-angle distortion distribution pattern, while for the latter, we derive rectification trajectories with backward optical flows and smooth them. Compared with ImagePC, VideoPC maintains high-quality facial corrections in space and mitigates the potential temporal shakes sequentially in blind scenarios. Finally, to establish an evaluation benchmark and train the framework, we establish a video portrait dataset with a large diversity in the number of people, lighting conditions, and background. Experiments demonstrate that the proposed methods outperform existing solutions quantitatively and qualitatively, contributing to high-fidelity wide-angle videos with stable and natural portraits.

Depth estimation from a monocular 360 image is important to the perception of the entire 3D environment. However, the inherent distortion and large field of view (FoV) in 360 images pose great challenges for this task. To this end, existing mainstream solutions typically introduce additional perspective-based 360 representations (e.g., Cubemap) to achieve effective feature extraction. Nevertheless, regardless of the introduced representations, they eventually need to be unified into the equirectangular projection (ERP) format for the subsequent depth estimation, which inevitably reintroduces additional distortions. In this work, we propose an oriented-distortion-aware Gabor Fusion framework (PGFuse) to address the above challenges. First, we introduce Gabor filters that analyze texture in the frequency domain, extending the receptive fields and enhancing depth cues. To address the reintroduced distortions, we design a latitude-aware distortion representation to generate customized, distortion-aware Gabor filters (PanoGabor filters). Furthermore, we design a channel-wise and spatial-wise unidirectional fusion module (CS-UFM) that integrates the proposed PanoGabor filters to unify other representations into the ERP format, delivering effective and distortion-aware features. Considering the orientation sensitivity of the Gabor transform, we further introduce a spherical gradient constraint to stabilize this sensitivity. Experimental results on three popular indoor 360 benchmarks demonstrate the superiority of the proposed PGFuse to existing state-of-the-art solutions. Code and models will be available at https://github.com/zhijieshen-bjtu/PGFuse.

Image stitching under large parallax remains a challenging task due to the conflict of content alignment and shape preservation. Most methods focus on precisely aligning overlapping regions via spatially varying transformations, often causing unexpected distortions in large-parallax areas. Differently, we aim to produce stitched images that are both visually natural and free of artifacts. To this end, we present a parallax-aware unsupervised warping model for seam-guided image stitching. To preserve natural content, we first design an edge-enhanced mask generation module to distinguish large-parallax regions and suppress excessive deformation around these areas. It is constrained by a comprehensive objective function that integrates masked photometric difference, nontrivial mask learning, and adaptive regularization, simultaneously ensuring mask reliability and alignment robustness. Besides, to eliminate parallax artifacts, we incorporate a seam-guided alignment strategy into our warping network, which iteratively registers local regions with the assistance of optimal seam estimation. Through adaptively finetuning the warping model, we progressively improve the stitching quality with improved seam quality. To facilitate the learning process of perceiving parallax, we construct a new image stitching dataset with larger parallax than that of UDISD, which could benefit the model’s generalization in challenging scenarios. Experiments show our solution not only removes misaligned regions but also maintains shape consistency especially in challenging parallax scenarios.

Multi-object tracking (MOT) in fisheye images becomes particularly challenging due to significant radial distortion. In MOT, Camera Motion Compensation (CMC) is crucial for mitigating inter-frame errors caused by camera movement. However, conventional CMC methods degrade in fisheye images, as their rigid motion models cannot characterize the non-uniform motion fields introduced by severe distortion. In this paper, we first establish this limitation through theoretical derivation and then propose a plug-and-play ROI-Centric Motion Compensation (RCMC) mechanism, which serves as a new CMC solution that leverages instance-aware ROI for motion estimation. Moreover, fisheye distortion introduces significant target tilt, distorting object geometry and thereby hindering accurate detection and separation of adjacent instances. To address this issue, we integrate RCMC with tailored adjustments for oriented bounding boxes (OBB) to form FO-MOT, a MOT framework for fisheye images. In addition, to overcome the lack of dedicated evaluation benchmarks, we present FisheyeMOT, a comprehensively annotated dataset with OBB-based labels, designed to support the development and validation of fisheye MOT systems. Finally, we demonstrate the superiority of our framework by qualitative and quantitative experiments. The dataset and code will be released at https://github.com/lukanightfever/FO-MOT.

Monocular 3D object detection is challenging due to the lack of accurate depth. However, existing depth-assisted solutions still exhibit inferior performance, whose reason is universally acknowledged as the unsatisfactory accuracy of monocular depth estimation models. In this paper, we revisit monocular 3D object detection from the depth perspective and formulate an additional issue as the limited 3D structure-aware capability of existing depth representations (e.g., depth one-hot encoding or depth distribution). To address this issue, we introduce a novel Depth Thickness Field approach to embed clear 3D structures of the scenes. Specifically, we present MonoDTF, a scene-to-instance depth-adapted network comprising a Scene-Level Depth Retargeting (SDR) module and an Instance-Level Spatial Refinement (ISR) module. The former retargets traditional depth representations to the proposed depth thickness field, incorporating the scene-level perception of 3D structures. The latter refines the voxel space with the guidance of instances, enhancing the 3D instance-aware capability of the depth thickness field and thus improving detection accuracy. Extensive experiments on the KITTI and Waymo datasets demonstrate our superiority to existing state-of-the-art (SoTA) methods and the universality when equipped with different depth estimation models. The source codes are available at https://github.com/QiuDeZhang/MonoDTF.

We present RopStitch, an unsupervised deep image stitching framework with both robustness and naturalness. To ensure the robustness of RopStitch, we propose to incorporate the universal prior of content perception into the image stitching model by a dual-branch architecture. It separately captures coarse and fine features and integrates them to achieve highly generalizable performance across diverse unseen real-world scenes. Concretely, the dual-branch model consists of a pretrained branch to capture semantically invariant representations and a learnable branch to extract fine-grained discriminative features, which are then merged into a whole by a controllable factor at the correlation level. Besides, considering that content alignment and structural preservation are often contradictory to each other, we propose a concept of virtual optimal planes to relieve this conflict. To this end, we model this problem as a process of estimating homography decomposition coefficients, and design an iterative coefficient predictor and minimal semantic distortion constraint to identify the optimal plane. This scheme is finally incorporated into RopStitch by warping both views onto the optimal plane bidirectionally. Extensive experiments across various datasets demonstrate that RopStitch significantly outperforms existing methods, particularly in scene robustness and content naturalness.

In this paper, we propose \textbf{Jasmine}, the first Stable Diffusion (SD)-based self-supervised framework for monocular depth estimation, which effectively harnesses SD’s visual priors to enhance the sharpness and generalization of unsupervised prediction. Previous SD-based methods are all supervised since adapting diffusion models for dense prediction requires high-precision supervision. In contrast, self-supervised reprojection suffers from inherent challenges (\textit{e.g.}, occlusions, texture-less regions, illumination variance), and the predictions exhibit blurs and artifacts that severely compromise SD's latent priors. To resolve this, we construct a novel surrogate task of mix-batch image reconstruction. Without any additional supervision, it preserves the detail priors of SD models by reconstructing the images themselves while preventing depth estimation from degradation. Furthermore, to address the inherent misalignment between SD's scale and shift invariant estimation and self-supervised scale-invariant depth estimation, we build the Scale-Shift GRU. It not only bridges this distribution gap but also isolates the fine-grained texture of SD output against the interference of reprojection loss. Extensive experiments demonstrate that Jasmine achieves SoTA performance on the KITTI benchmark and exhibits superior zero-shot generalization across multiple datasets.

We retarget video stitching to an emerging issue, named warping shake, which unveils the temporal content shakes induced by sequentially unsmooth warps when extending image stitching to video stitching. Even if the input videos are stable, the stitched video can inevitably cause undesired warping shakes and affect the visual experience. To address this issue, we propose StabStitch++, a novel video stitching framework to realize spatial stitching and temporal stabilization with unsupervised learning simultaneously. First, different from existing learning-based image stitching solutions that typically warp one image to align with another, we suppose a virtual midplane between original image planes and project them onto it. Concretely, we design a differentiable bidirectional decomposition module to disentangle the homography transformation and incorporate it into our spatial warp, evenly spreading alignment burdens and projective distortions across two views. Then, inspired by camera paths in video stabilization, we derive the mathematical expression of stitching trajectories in video stitching by elaborately integrating spatial and temporal warps. Finally, a warp smoothing model is presented to produce stable stitched videos with a hybrid loss to simultaneously encourage content alignment, trajectory smoothness, and online collaboration. Compared with StabStitch that sacrifices alignment for stabilization, StabStitch++ makes no compromise and optimizes both of them simultaneously, especially in the online mode. To establish an evaluation benchmark and train the learning framework, we build a video stitching dataset with a rich diversity in camera motions and scenes. Experiments exhibit that StabStitch++ surpasses current solutions in stitching performance, robustness, and efficiency, offering compelling advancements in this field by building a real-time online video stitching system.

Panoramic distortion poses a significant challenge in 360° depth estimation, particularly pronounced at the north and south poles. Existing methods either adopt a bi-projection fusion strategy to remove distortions or model long-range dependencies to capture global structures, resulting in either unclear structure or insufficient local perception. In this paper, we propose a spherical geometry transformer, named SGFormer, to address the above issues, with an innovative step to integrate spherical geometric priors into vision transformers. To this end, we retarget the transformer decoder to a spherical prior decoder (termed SPDecoder), which endeavors to uphold the integrity of spherical structures during decoding. Concretely, we leverage bipolar reprojection, circular rotation, and curve local embedding to preserve the spherical characteristics of equidistortion, continuity, and surface distance, respectively. Furthermore, we present a query-based global conditional position embedding to compensate for spatial structure at varying resolutions. It not only boosts the global perception of spatial position but also sharpens the depth structure across different patches. Finally, we conduct extensive experiments on popular benchmarks, demonstrating our superiority over state-of-the-art solutions. Our code will be made publicly athttps://github.com/iuiuJaon/SGFormer.

Iterative inference approaches have shown promising success in the task of multi-view depth estimation. However, these methods put excessive emphasis on the universal inter-view correspondences while neglecting the correspondence ambiguity in regions of low texture and depth discontinuous areas. Thus, they are prone to produce inaccurate or even erroneous depth estimations, which is further exacerbated due to cumulative errors especially in the iterative pipeline, providing unreliable information in many real-world scenarios. In this article, we revisit this issue from the intra-view contextual hints and introduce a novel enhancing iterative approach, named EnIter. Concretely, at the beginning of each iteration, we present a Depth Intercept (DI) modulator to provide more accurate depth by aggregating neighbor uncertainty, correlation volume of reference and normal. This plug and play modulator is effective at intercepting the erroneous depth estimations with implicit guidance from the universal correlation contextual hints, especially for the challenging regions. Furthermore, at the end of each iteration, we refine the depth map with another plug and play modulator termed as Depth Refine (DR). It mines the latent structure knowledge of reference contextual hints and establishes one-way dependency using local attention from reference features to depth, yielding delicate depth in detail. Extensive experiment demonstrates that our method not only achieves state-of-the-art performance over existing models but also exhibits remarkable universality in popular iterative pipelines, e.g., CasMVS, UCSNet, TransMVS, and UniMVS.

Fisheye object detection is challenging due to the fisheye distortion, which inclines objects to different extents and pushes extensive irrelevant pixels into the predicted horizontal bounding box (HBB). To address the problems above, we establish a new fisheye object detection dataset (named FishOBB) with compact oriented bounding box (OBB) annotations, as well as an OBB-customized mosaic augmentation technology. To our knowledge, there are very few fisheye datasets labeled by OBB, especially the open source forword view dataset like ours. Besides, we provide a fisheye object detection baseline (named FDA-YOLO) with two fisheye adaption units. Concretely, we first design a distortion orientation aggregation (DOA) unit guided by polar sampling to capture distortion-aware fisheye features. On the other hand, to transfer HBB-based detection models to OBB-based counterparts, we propose an oriented anchor attention unit. It automatically weights the unbalanced positive/negative samples and facilitates convergence for multi-anchor models. Finally, we demonstrate that the two adaption units can be easily integrated into various anchor-based YOLO methods, e.g., ScaledYOLOv4 and YOLOv7, contributing to superior performance to existing state-of-the-art (SoTA) solutions in the proposed dataset. Meanwhile, our method has also achieved SoTA performance on other popular datasets like WEPDTOF. The dataset and code are released at https://github.com/lukanightfever/FishOBB .

Depth estimation models have shown promising performance on clear scenes but fail to generalize to adverse weather conditions due to illumination variations, weather particles, etc. In this paper, we propose WeatherDepth, a self-supervised robust depth estimation model with curriculum contrastive learning, to tackle performance degradation in complex weather conditions. Concretely, we first present a progressive curriculum learning scheme with three simple-to-complex curricula to gradually adapt the model from clear to relative adverse, and then to adverse weather scenes. It encourages the model to gradually grasp beneficial depth cues against the weather effect, yielding smoother and better domain adaption. Meanwhile, to prevent the model from forgetting previous curricula, we integrate contrastive learning into different curricula. By drawing reference knowledge from the previous course, our strategy establishes a depth consistency constraint between different courses toward robust depth estimation in diverse weather. Besides, to reduce manual intervention and better adapt to different models, we designed an adaptive curriculum scheduler to automatically search for the best timing for course switching. In the experiment, the proposed solution is proven to be easily incorporated into various architectures and demonstrates state-of-the-art (SoTA) performance on both synthetic and real weather datasets. Source code and data are available at https://github.com/wangjiyuan9/WeatherDepth.

Recently, diffusion-based depth estimation methods have drawn widespread attention due to their elegant denoising patterns and promising performance. However, they are typically unreliable under adverse conditions prevalent in real-world scenarios, such as rainy, snowy, etc. In this paper, we propose a novel robust depth estimation method called D4RD, featuring a custom contrastive learning mode tailored for diffusion models to mitigate performance degradation in complex environments. Concretely, we integrate the strength of knowledge distillation into contrastive learning, building the `trinity' contrastive scheme. This scheme utilizes the sampled noise of the forward diffusion process as a natural reference, guiding the predicted noise in diverse scenes toward a more stable and precise optimum. Moreover, we extend noise-level trinity to encompass more generic feature and image levels, establishing a multi-level contrast to distribute the burden of robust perception across the overall network. Before addressing complex scenarios, we enhance the stability of the baseline diffusion model with three straightforward yet effective improvements, which facilitate convergence and remove depth outliers. Extensive experiments demonstrate that D4RD surpasses existing state-of-the-art solutions on synthetic corruption datasets and real-world weather conditions. Source code and data are available at \url{https://github.com/wangjiyuan9/D4RD}.

Irregular boundaries in image stitching naturally occur due to freely moving cameras. To deal with this problem, existing methods focus on optimizing mesh warping to make boundaries regular using the traditional explicit solution. However, previous methods always depend on hand-crafted features (e.g., keypoints and line segments). Thus, failures often happen in overlapping regions without distinctive features. In this paper, we address this problem by proposing RecStitchNet, a reasonable and effective network for image stitching with rectangular boundaries. Considering that both stitching and imposing rectangularity are non-trivial tasks in the learning-based framework, we propose a three-step progressive learning based strategy, which not only simplifies this task, but gradually achieves a good balance between stitching and imposing rectangularity. In the first step, we perform initial stitching by a pre-trained state-of-the-art image stitching model, to produce initially warped stitching results without considering the boundary constraint. Then, we use a regression network with a comprehensive objective regarding mesh, perception, and shape to further encourage the stitched meshes to have rectangular boundaries with high content fidelity. Finally, we propose an unsupervised instance-wise optimization strategy to refine the stitched meshes iteratively, which can effectively improve the stitching results in terms of feature alignment, as well as boundary and structure preservation. Due to the lack of stitching datasets and the difficulty of label generation, we propose to generate a stitching dataset with rectangular stitched images as pseudo-ground-truth labels, and the performance upper bound induced from the it can be broken by our unsupervised refinement. Qualitative and quantitative results and evaluations demonstrate the advantages of our method over the state-of-the-art.

Thin-plate spline (TPS) is a principal warp that allows for representing elastic, nonlinear transformation with control point motions. With the increase of control points, the warp becomes increasingly flexible but usually encounters a bottleneck caused by undesired issues, e.g., content distortion. In this paper, we explore generic applications of TPS in single-image-based warping tasks, such as rotation correction, rectangling, and portrait correction. To break this bottleneck, we propose the coupled thin-plate spline model (CoupledTPS), which iteratively couples multiple TPS with limited control points into a more flexible and powerful transformation. Concretely, we first design an iterative search to predict new control points according to the current latent condition. Then, we present the warping flow as a bridge for the coupling of different TPS transformations, effectively eliminating interpolation errors caused by multiple warps. Besides, in light of the laborious annotation cost, we develop a semi-supervised learning scheme to improve warping quality by exploiting unlabeled data. It is formulated through dual transformation between the searched control points of unlabeled data and its graphic augmentation, yielding an implicit correction consistency constraint. Finally, we collect massive unlabeled data to exhibit the benefit of our semi-supervised scheme in rotation correction. Extensive experiments demonstrate the superiority and universality of CoupledTPS over the existing State-of-the-Art (SoTA) solutions for rotation correction and beyond.

Existing panoramic layout estimation solutions tend to recover room boundaries from a vertically compressed sequence, yielding imprecise results as the compression process often muddles the semantics between various planes. Besides, these data-driven approaches impose an urgent demand for massive data annotations, which are laborious and time-consuming. For the first problem, we propose an orthogonal plane disentanglement network (termed DOPNet) to distinguish ambiguous semantics. DOPNet consists of three modules that are integrated to deliver distortion-free, semantics-clean, and detail-sharp disentangled representations, which benefit the subsequent layout recovery. For the second problem, we present an unsupervised adaptation technique tailored for horizon-depth and ratio representations. Concretely, we introduce an optimization strategy for decision-level layout analysis and a 1D cost volume construction method for feature-level multi-view aggregation, both of which are designed to fully exploit the geometric consistency across multiple perspectives. The optimizer provides a reliable set of pseudo-labels for network training, while the 1D cost volume enriches each view with comprehensive scene information derived from other perspectives. Extensive experiments demonstrate that our solution outperforms other SoTA models on both monocular layout estimation and multi-view layout estimation tasks.

Category-level 6D object pose estimation aims to estimate the pose and size of unseen objects with known categories. Existing methods mainly focus on capturing geometric features to handle shape variations, and are prone to failure in occlusion and noisy environments. In this paper, we propose TG-Pose, a unified pose estimation framework that delves into topology and geometry to deal with the above issues. To exploit topological properties, we first propose a topological feature predictor and a topological label generator to dig into the underlying structural details from encoded features using persistent homology. Then, the topological and geometric features are employed to facilitate the symmetry reconstruction of the original point cloud to obtain a reliable and coherent object shape, which, in turn, guides the pose estimation. For each object category, we construct geometric and topological templates by leveraging inherent intra-class similarities. These templates enhance the reliability of pose estimation and the completeness of object structure through geometric alignment and topological guidance, especially when handling incomplete objects. Moreover, a pose-aware enhancement strategy is designed to enhance the encoder in learning pose-sensitive features and robustness to noisy point clouds. Experimental results show that TG-Pose outperforms the state-of-the-art solutions on public benchmarks and achieves better generalization in real-world datasets. Project Page https://sites.google.com/view/tg-pose.

Based on the Manhattan World assumption, most existing indoor layout estimation schemes focus on recovering layouts from vertically compressed 1D sequences. However, the compression procedure confuses the semantics of different planes, yielding inferior performance with ambiguous interpretability. To address this issue, we propose to disentangle this 1D representation by pre-segmenting orthogonal (vertical and horizontal) planes from a complex scene, explicitly capturing the geometric cues for indoor layout estimation. Considering the symmetry between the floor boundary and ceiling boundary, we also design a soft-flipping fusion strategy to assist the pre-segmentation. Besides, we present a feature assembling mechanism to effectively integrate shallow and deep features with distortion distribution awareness. To compensate for the potential errors in pre-segmentation, we further leverage triple attention to reconstruct the disentangled sequences for better performance. Experiments on four popular benchmarks demonstrate our superiority over existing SoTA solutions, especially on the 3DIoU metric. The code is available at https://github.com/zhijieshen-bjtu/DOPNet.

The wide-angle lens shows appealing applications in VR technologies, but it introduces severe radial distortion into its captured image. To recover the realistic scene, previous works devote to rectifying the content of the wide-angle image. However, such a rectification solution inevitably distorts the image boundary, which changes related geometric distributions and misleads the current vision perception models. In this work, we explore constructing a win-win representation on both content and boundary by contributing a new learning model, i.e., Rectangling Rectification Network (RecRecNet). In particular, we propose a thin-plate spline (TPS) module to formulate the nonlinear and non-rigid transformation for rectangling images. By learning the control points on the rectified image, our model can flexibly warp the source structure to the target domain and achieves an end-to-end unsupervised deformation. To relieve the complexity of structure approximation, we then inspire our RecRecNet to learn the gradual deformation rules with a DoF (Degree of Freedom)-based curriculum learning. By increasing the DoF in each curriculum stage, namely, from similarity transformation (4-DoF) to homography transformation (8-DoF), the network is capable of investigating more detailed deformations, offering fast convergence on the final rectangling task. Experiments show the superiority of our solution over the compared methods on both quantitative and qualitative evaluations. The code and dataset are available at https://github.com/KangLiao929/RecRecNet.

Optical flow, or the estimation of motion fields from image sequences, is one of the fundamental problems in computer vision. Unlike most pixel-wise tasks that aim at achieving consistent representations of the same category, optical flow raises extra demands for obtaining local discrimination and smoothness, which yet is not fully explored by existing approaches. In this paper, we push Gaussian Attention (GA) into the optical flow models to accentuate local properties during representation learning and enforce the motion affinity during matching. Specifically, we introduce a novel Gaussian-Constrained Layer (GCL) which can be easily plugged into existing Transformer blocks to highlight the local neighborhood that contains fine-grained structural information. Moreover, for reliable motion analysis, we provide a new Gaussian-Guided Attention Module (GGAM) which not only inherits properties from Gaussian distribution to instinctively revolve around the neighbor fields of each point but also is empowered to put the emphasis on contextually related regions during matching. Our fully-equipped model, namely Gaussian Attention Flow network (GAFlow), naturally incorporates a series of novel Gaussian-based modules into the conventional optical flow framework for reliable motion analysis. Extensive experiments on standard optical flow datasets consistently demonstrate the exceptional performance of the proposed approach in terms of both generalization ability evaluation and online benchmark testing. Code is available at https://github.com/LA30/GAFlow.

Traditional image stitching approaches tend to leverage increasingly complex geometric features (e.g., point, line, edge, etc.) for better performance. However, these hand-crafted features are only suitable for specific natural scenes with adequate geometric structures. In contrast, deep stitching schemes overcome adverse conditions by adaptively learning robust semantic features, but they cannot handle large-parallax cases.To solve these issues, we propose a parallax-tolerant unsupervised deep image stitching technique. First, we propose a robust and flexible warp to model the image registration from global homography to local thin-plate spline motion. It provides accurate alignment for overlapping regions and shape preservation for non-overlapping regions by joint optimization concerning alignment and distortion. Subsequently, to improve the generalization capability, we design a simple but effective iterative strategy to enhance the warp adaption in cross-dataset and cross-resolution applications. Finally, to further eliminate the parallax artifacts, we propose to composite the stitched image seamlessly by unsupervised learning for seam-driven composition masks. Compared with existing methods, our solution is parallax-tolerant and free from laborious designs of complicated geometric features for specific scenes. Extensive experiments show our superiority over the SoTA methods, both quantitatively and qualitatively. The code is available at https://github.com/nie-lang/UDIS2.

Omnidirectional multi-view stereo (MVS) vision is attractive for its ultra-wide field-of-view (FoV), enabling machines to perceive 360°3D surroundings. However, the existing solutions require expensive dense depth labels for supervision, making them impractical in real-world applications. In this paper, we propose the first unsupervised omnidirectional MVS framework based on multiple fisheye images. To this end, we project all images to a virtual view center and composite two panoramic images with spherical geometry from two pairs of back-to-back fisheye images. The two 360° images formulate a stereo pair with a special pose, and the photometric consistency is leveraged to establish the unsupervised constraint, which we term “Pseudo-Stereo Supervision”. In addition, we propose Un-OmniMVS, an efficient unsupervised omnidirectional MVS network, to facilitate the inference speed with two efficient components. First, a novel feature extractor with frequency attention is proposed to simultaneously capture the non-local Fourier features and local spatial features, explicitly facilitating the feature representation. Then, a variance-based light cost volume is put forward to reduce the computational complexity. Experiments exhibit that the performance of our unsupervised solution is competitive to that of the state-of-the-art (SoTA) supervised methods with better generalization in real-world data. The code will be available at https://github.com/Chen-z-s/Un-OmniMVS.

Multi-fisheye stereo matching is a promising task that employs the traditional multi-view stereo (MVS) pipeline with spherical sweeping to acquire omnidirectional depth. However, the existing omnidirectional MVS technologies neglect fisheye and omnidirectional distortions, yielding inferior performance. In this paper, we revisit omnidirectional MVS by incorporating three sphere geometry priors: spherical projection, spherical continuity, and spherical position. To deal with fisheye distortion, we propose a new distortion-adaptive fusion module to convert fisheye inputs into distortion-free spherical tangent representations by constructing a spherical projection space. Then these multi-scale features are adaptively aggregated with additional learnable offsets to enhance content perception. To handle omnidirectional distortion, we present a new spherical cost aggregation module with a comprehensive consideration of the spherical continuity and position. Concretely, we first design a rotation continuity compensation mechanism to ensure omnidirectional depth consistency of left-right boundaries without introducing extra computation. On the other hand, we encode the geometry-aware spherical position and push them into the cost aggregation to relieve panoramic distortion and perceive the 3D structure. Furthermore, to avoid the excessive concentration of depth hypothesis caused by inverse depth linear sampling, we develop a segmented sampling strategy that combines linear and exponential spaces to create S-OmniMVS, along with three sphere priors. Extensive experiments demonstrate the proposed method outperforms the state-of-the-art (SoTA) solutions by a large margin on various datasets both quantitatively and qualitatively.

Semantic segmentation on 360° images is a vital component of scene understanding due to the rich surrounding information. Recently, horizontal representation-based approaches outperform projection-based solutions, because the distortions can be effectively removed by compressing the spherical data in the vertical direction. However, these methods ignore the distortion distribution prior and are limited to unbalanced receptive fields, e.g., the receptive fields are sufficient in the vertical direction and insufficient in the horizontal direction. Differently, a vertical representation compressed in another direction can offer implicit distortion prior and enlarge horizontal receptive fields. In this paper, we combine the two different representations and propose a novel 360° semantic segmentation solution from a complementary perspective. Our network comprises three modules: a feature extraction module, a bi-directional compression module, and an ensemble decoding module. First, we extract multi-scale features from a panorama. Then, a bi-directional compression module is designed to compress features into two complementary low-dimensional representations, which provide content perception and distortion prior. Furthermore, to facilitate the fusion of bi-directional features, we design a unique self distillation strategy in the ensemble decoding module to enhance the interaction of different features and further improve the performance. Experimental results show that our approach outperforms the state-of-the-art solutions on quantitative evaluations while displaying the best performance on visual appearance.

Depth-image-based rendering (DIBR) technologies have been widely employed to synthesize novel realistic views from a single image in 3D video applications. However, DIBR-oriented approaches heavily rely on the accuracy of depth maps, usually requiring the depth GT as a prior. Despite that, there might exist extensive float precision losses and invalid holes in the synthesized view due to warping error and occlusion. In this paper, we propose an end-to-end as-deformable-as-possible (ADAP) single-image-based view synthesis solution without depth prior. It addresses the above issues through two stages: alignment and reconstruction, where we first transform the input image to the latent feature space and then reconstruct the novel view in the image domain. In the first stage, the input image is deformed to align with the synthesized view at feature level. To this end, we propose an ADAP alignment mechanism through pixel-level warping to error-level quantization to feature-level alignment, progressively improving the deformable capability in handling challenging motion conditions in real-world scenes. In the second stage, we exploit an occlusion-aware reconstruction module to recover the content details from the deformed feature at pixel level. Extensive experiments demonstrate that our alignment-reconstruction approach is robust to the depth map. Even with a coarsely estimated depth map, our solution outperforms other SoTA schemes in the popular benchmarks.

Not everybody can be equipped with professional photography skills and sufficient shooting time, and there can be some tilts in the captured images occasionally. In this paper, we propose a new and practical task, named Rotation Correction, to automatically correct the tilt with high content fidelity in the condition that the rotated angle is unknown. This task can be easily integrated into image editing applications, allowing users to correct the rotated images without any manual operations. To this end, we leverage a neural network to predict the optical flows that can warp the tilted images to be perceptually horizontal. Nevertheless, the pixel-wise optical flow estimation from a single image is severely unstable, especially in large-angle tilted images. To enhance its robustness, we propose a simple but effective prediction strategy to form a robust elastic warp. Particularly, we first regress the mesh deformation that can be transformed into robust initial optical flows. Then we estimate residual optical flows to facilitate our network the flexibility of pixel-wise deformation, further correcting the details of the tilted images. To establish an evaluation benchmark and train the learning framework, a comprehensive rotation correction dataset is presented with a large diversity in scenes and rotated angles. Extensive experiments demonstrate that even in the absence of the angle prior, our algorithm can outperform other state-of-the-art solutions requiring this prior. The code and dataset are available at https://github.com/nie-lang/RotationCorrection.

Stitched images provide a wide field-of-view (FoV) but suffer from unpleasant irregular boundaries. To deal with this problem, existing image rectangling methods devote to searching an initial mesh and optimizing a target mesh to form the mesh deformation in two stages. Then rectangu-lar images can be generated by warping stitched images. However, these solutions only work for images with rich linear structures, leading to noticeable distortions for por-traits and landscapes with non-linear objects. In this paper, we address these issues by proposing the first deep learning solution to image rectangling. Con-cretely, we predefine a rigid target mesh and only estimate an initial mesh to form the mesh deformation, contributing to a compact one-stage solution. The initial mesh is predicted using a fully convolutional network with a resid-ual progressive regression strategy. To obtain results with high content fidelity, a comprehensive objective function is proposed to simultaneously encourage the boundary rect-angular, mesh shape-preserving, and content perceptually natural. Besides, we build the first image stitching rectan-gling dataset with a large diversity in irregular boundaries and scenes. Experiments demonstrate our superiority over traditional methods both quantitatively and qualitatively.

Single-image-based view synthesis is significant for generating a 3D scene and gains increasing attention in recent years. However, this task is challenging as it requires inferring contents beyond what is immediately visible. Previous methods directly predict the unknown views using the convolutional neural networks, but the generated views suffer from visually unpleasant holes, deformations, and artifacts. In this paper, we propose a Single-image-based view synthesis transformer (named SivsFormer) for high-quality and realistic view synthesis. In particular, a warping and occlusion handing module is designed to reduce the influence of parallax on the network. Subsequently, a disparity alignment module captures the long-range information over the scene and ensures that pixels move in a geometrically correct manner with soft probabilistic disparity maps. Moreover, we present a parallax-aware loss function to improve the quality of the synthetic images, which explicitly quantifies the magnitude of parallaxes. We conduct extensive experiments on popular KITTI and Cityscapes datasets. Benefitting from the proposed parallax-aware transformer, our approach achieves superior performance in both quantitative and qualitative evaluations.

Homography estimation is an important task in computer vision applications, such as image stitching, video stabilization, and camera calibration. Traditional homography estimation methods heavily depend on the quantity and distribution of feature correspondences, leading to poor robustness in low-texture scenes. The learning solutions, on the contrary, try to learn robust deep features but demonstrate unsatisfying performance in the scenes with low overlap rates. In this paper, we address these two problems simultaneously by designing a contextual correlation layer (CCL). The CCL can efficiently capture the long-range correlation within feature maps and can be flexibly used in a learning framework. In addition, considering that a single homography can not represent the complex spatial transformation in depth-varying images with parallax, we propose to predict multi-grid homography from global to local. Moreover, we equip our network with a depth perception capability, by introducing a novel depth-aware shape-preserved loss. Extensive experiments demonstrate the superiority of our method over state-of-the-art solutions in the synthetic benchmark dataset and real-world dataset. The codes and models will be available athttps://github.com/nie-lang/Multi-Grid-Deep-Homography.

For a monocular 360° image, depth estimation is a challenging because the distortion increases along the latitude. To perceive the distortion, existing methods devote to designing a deep and complex network architecture. In this paper, we provide a new perspective that constructs an interpretable and sparse representation for a 360° image. Considering the importance of the geometric structure in depth estimation, we utilize the contourlet transform to capture an explicit geometric cue in the spectral domain and integrate it with an implicit cue in the spatial domain. Specifically, we propose a neural contourlet network consisting of a convolutional neural network and a contourlet transform branch. In the encoder stage, we design a spatial–spectral fusion module to effectively fuse two types of cues. Contrary to the encoder, we employ the inverse contourlet transform with learned low-pass subbands and band-pass directional subbands to compose the depth in the decoder. Experiments on the three popular 360° panoramic image datasets demonstrate that the proposed approach outperforms the state-of-the-art schemes with faster convergence. Code is available athttps://github.com/zhijieshen-bjtu/Neural-Contourlet-Network-for-MODE.

Estimating the depth of omnidirectional images is more challenging than that of normal field-of-view (NFoV) images because the varying distortion can significantly twist an object’s shape. The existing methods suffer from troublesome distortion while estimating the depth of omnidirectional images, leading to inferior performance. To reduce the negative impact of the distortion influence, we propose a distortion-tolerant omnidirectional depth estimation algorithm using a dual-cubemap. It comprises two modules: Dual-Cubemap Depth Estimation (DCDE) module and Boundary Revision (BR) module. In DCDE module, we present a rotation-based dual-cubemap model to estimate the accurate NFoV depth, reducing the distortion at the cost of boundary discontinuity on omnidirectional depths. Then a boundary revision module is designed to smooth the discontinuous boundaries, which contributes to the precise and visually continuous omnidirectional depths. Extensive experiments demonstrate the superiority of our method over other state-of-the-art solutions.

Traditional feature-based image stitching technologies rely heavily on feature detection quality, often failing to stitch images with few features or low resolution. The learning-based image stitching solutions are rarely studied due to the lack of labeled data, making the supervised methods unreliable. To address the above limitations, we propose an unsupervised deep image stitching framework consisting of two stages: unsupervised coarse image alignment and unsupervised image reconstruction. In the first stage, we design an ablation-based loss to constrain an unsupervised homography network, which is more suitable for large-baseline scenes. Moreover, a transformer layer is introduced to warp the input images in the stitching-domain space. In the second stage, motivated by the insight that the misalignments in pixel-level can be eliminated to a certain extent in feature-level, we design an unsupervised image reconstruction network to eliminate the artifacts from features to pixels. Specifically, the reconstruction network can be implemented by a low-resolution deformation branch and a high-resolution refined branch, learning the deformation rules of image stitching and enhancing the resolution simultaneously. To establish an evaluation benchmark and train the learning framework, a comprehensive real-world image dataset for unsupervised deep image stitching is presented and released. Extensive experiments well demonstrate the superiority of our method over other state-of-the-art solutions. Even compared with the supervised solutions, our image stitching quality is still preferred by users.