Vision Transformers have achieved impressive performance in video classification, while suffering from the quadratic complexity caused by the Softmax attention mechanism. Some studies alleviate the computational costs by reducing the number of tokens in attention calculation, but the complexity is still quadratic. Another promising way is to replace Softmax attention with linear attention, which owns linear complexity but presents a clear performance drop. We find that such a drop in linear attention results from the lack of attention concentration on critical features. Therefore, we propose a feature fixation module to reweight the feature importance of the query and key before computing linear attention. Specifically, we regard the query, key, and value as various latent representations of the input token, and learn the feature fixation ratio by aggregating Query-Key-Value information. This is beneficial for measuring the feature importance comprehensively. Furthermore, we enhance the feature fixation by neighborhood association, which leverages additional guidance from spatial and temporal neighbouring tokens. The proposed method significantly improves the linear attention baseline and achieves state-of-the-art performance among linear video Transformers on three popular video classification benchmarks. With fewer parameters and higher efficiency, our performance is even comparable to some Softmax-based quadratic Transformers.
A single rolling-shutter (RS) image may be viewed as a row-wise combination of a sequence of global-shutter (GS) images captured by a (virtual) moving GS camera within the exposure duration. Although RS cameras are widely used, the RS effect causes obvious image distortion especially in the presence of fast camera motion, hindering downstream computer vision tasks. In this paper, we propose to invert the RS image capture mechanism, i.e., recovering a continuous high framerate GS video from two time-consecutive RS frames. We call this task the RS temporal super-resolution (RSSR) problem. The RSSR is a very challenging task, and to our knowledge, no practical solution exists to date. This paper presents a novel deep-learning based solution. By leveraging the multi-view geometry relationship of the RS imaging process, our learning-based framework successfully achieves high framerate GS generation. Specifically, three novel contributions can be identified: (i) novel formulations for bidirectional RS undistortion flows under constant velocity as well as constant acceleration motion model. (ii) a simple linear scaling operation, which bridges the RS undistortion flow and regular optical flow. (iii) a new mutual conversion scheme between varying RS undistortion flows that correspond to different scanlines. Our method also exploits the underlying spatial-temporal geometric relationships within a deep learning framework, where no additional supervision is required beyond the necessary middle-scanline GS image. Building upon these contributions, we represent the very first rolling-shutter temporal super-resolution deep-network that is able to recover high framerate GS videos from just two RS frames. Extensive experimental results on both synthetic and real data show that our proposed method can produce high-quality GS image sequences with rich details, outperforming the state-of-the-art methods.
Accurate moving object segmentation is an essential task for autonomous driving. It can provide effective information for many downstream tasks, such as collision avoidance, path planning, and static map construction. How to effectively exploit the spatial-temporal information is a critical question for 3D LiDAR moving object segmentation (LiDAR-MOS). In this work, we propose a novel deep neural network exploiting both spatial-temporal information and different representation modalities of LiDAR scans to improve LiDAR-MOS performance. Specifically, we first use a range image-based dual-branch structure to separately deal with spatial and temporal information that can be obtained from sequential LiDAR scans, and later combine them using motion-guided attention modules. We also use a point refinement module via 3D sparse convolution to fuse the information from both LiDAR range image and point cloud representations and reduce the artifacts on the borders of the objects. We verify the effectiveness of our proposed approach on the LiDAR-MOS benchmark of SemanticKITTI. Our method outperforms the state-of-the-art methods significantly in terms of LiDAR-MOS IoU. Benefiting from the devised coarse-to-fine architecture, our method operates online at sensor frame rate. The implementation of our method is available as open source at: https://github.com/haomo-ai/MotionSeg3D.
Recently, Neural Radiance Fields (NeRF) is revolutionizing the task of novel view synthesis (NVS) for its superior performance. However, NeRF and its variants generally require a lengthy per-scene training procedure, where a multi-layer perceptron (MLP) is fitted to the captured images. To remedy the challenge, the voxel-grid representation has been proposed to significantly speed up the training. However, these existing methods can only deal with static scenes. How to develop an efficient and accurate dynamic view synthesis method remains an open problem. Extending the methods for static scenes to dynamic scenes is not straightforward as both the scene geometry and appearance change over time. In this paper, built on top of the recent advances in voxel-grid optimization, we propose a fast deformable radiance field method to handle dynamic scenes. Our method consists of two modules. The first module adopts a deformation grid to store 3D dynamic features, and a light-weight MLP for decoding the deformation that maps a 3D point in observation space to the canonical space using the interpolated features. The second module contains a density and a color grid to model the geometry and density of the scene. The occlusion is explicitly modeled to further improve the rendering quality. Experimental results show that our method achieves comparable performance to D-NeRF using only 20 minutes for training, which is more than 70x faster than D-NeRF, clearly demonstrating the efficiency of our proposed method.
With the ubiquity of rolling shutter (RS) cameras, it is becoming increasingly attractive to recover the latent global shutter (GS) video from two consecutive RS frames, which also places a higher demand on realism. Existing solutions, using deep neural networks or optimization, achieve promising performance. However, these methods generate intermediate GS frames through image warping based on the RS model, which inevitably result in black holes and noticeable motion artifacts. In this paper, we alleviate these issues by proposing a context-aware GS video reconstruction architecture. It facilitates the advantages such as occlusion reasoning, motion compensation, and temporal abstraction. Specifically, we first estimate the bilateral motion field so that the pixels of the two RS frames are warped to a common GS frame accordingly. Then, a refinement scheme is proposed to guide the GS frame synthesis along with bilateral occlusion masks to produce high-fidelity GS video frames at arbitrary times. Furthermore, we derive an approximated bilateral motion field model, which can serve as an alternative to provide a simple but effective GS frame initialization for related tasks. Experiments on synthetic and real data show that our approach achieves superior performance over state-of-the-art methods in terms of objective metrics and subjective visual quality. Code is available at \url{https://github.com/GitCVfb/CVR}.
Preys in the wild evolve to be camouflaged to avoid being recognized by predators. In this way, camouflage acts as a key defence mechanism across species that is critical to survival. To detect and segment the whole scope of a camouflaged object, camouflaged object detection (COD) is introduced as a binary segmentation task, with the binary ground truth camouflage map indicating the exact regions of the camouflaged objects. In this paper, we revisit this task and argue that the binary segmentation setting fails to fully understand the concept of camouflage. We find that explicitly modeling the conspicuousness of camouflaged objects against their particular backgrounds can not only lead to a better understanding about camouflage, but also provide guidance to designing more sophisticated camouflage techniques. Furthermore, we observe that it is some specific parts of camouflaged objects that make them detectable by predators. With the above understanding about camouflaged objects, we present the first triple-task learning framework to simultaneously localize, segment and rank camouflaged objects, indicating the conspicuousness level of camouflage. As no corresponding datasets exist for either the localization model or the ranking model, we generate localization maps with an eye tracker, which are then processed according to the instance level labels to generate our ranking-based training and testing dataset. We also contribute the largest COD testing set to comprehensively analyse performance of the camouflaged object detection models. Experimental results show that our triple-task learning framework achieves new state-of-the-art, leading to a more explainable camouflaged object detection network. Our code, data and results are available at: https://github.com/JingZhang617/COD-Rank-Localize-and-Segment.
Directly regressing the non-rigid shape and camera pose from the individual 2D frame is ill-suited to the Non-Rigid Structure-from-Motion (NRSfM) problem. This frame-by-frame 3D reconstruction pipeline overlooks the inherent spatial-temporal nature of NRSfM, i.e., reconstructing the whole 3D sequence from the input 2D sequence. In this paper, we propose to model deep NRSfM from a sequence-to-sequence translation perspective, where the input 2D frame sequence is taken as a whole to reconstruct the deforming 3D non-rigid shape sequence. First, we apply a shape-motion predictor to estimate the initial non-rigid shape and camera motion from a single frame. Then we propose a context modeling module to model camera motions and complex non-rigid shapes. To tackle the difficulty in enforcing the global structure constraint within the deep framework, we propose to impose the union-of-subspace structure by replacing the self-expressiveness layer with multi-head attention and delayed regularizers, which enables end-to-end batch-wise training. Experimental results across different datasets such as Human3.6M, CMU Mocap and InterHand prove the superiority of our framework. The code will be made publicly available
3D point cloud registration is fragile to outliers, which are labeled as the points without corresponding points. To handle this problem, a widely adopted strategy is to estimate the relative pose based only on some accurate correspondences, which is achieved by building correspondences on the identified inliers or by selecting reliable ones. However, these approaches are usually complicated and time-consuming. By contrast, the virtual point-based methods learn the virtual corresponding points (VCPs) for all source points uniformly without distinguishing the outliers and the inliers. Although this strategy is time-efficient, the learned VCPs usually exhibit serious collapse degeneration due to insufficient supervision and the inherent distribution limitation. In this paper, we propose to exploit the best of both worlds and present a novel robust 3D point cloud registration framework. We follow the idea of the virtual point-based methods but learn a new type of virtual points called rectified virtual corresponding points (RCPs), which are defined as the point set with the same shape as the source and with the same pose as the target. Hence, a pair of consistent point clouds, i.e. source and RCPs, is formed by rectifying VCPs to RCPs (VRNet), through which reliable correspondences between source and RCPs can be accurately obtained. Since the relative pose between source and RCPs is the same as the relative pose between source and target, the input point clouds can be registered naturally. Specifically, we first construct the initial VCPs by using an estimated soft matching matrix to perform a weighted average on the target points. Then, we design a correction-walk module to learn an offset to rectify VCPs to RCPs, which effectively breaks the distribution limitation of VCPs. Finally, we develop a hybrid loss function to enforce the shape and geometry structure consistency ...
3D point cloud registration in remote sensing field has been greatly advanced by deep learning based methods, where the rigid transformation is either directly regressed from the two point clouds (correspondences-free approaches) or computed from the learned correspondences (correspondences-based approaches). Existing correspondences-free methods generally learn the holistic representation of the entire point cloud, which is fragile for partial and noisy point clouds. In this paper, we propose a correspondences-free unsupervised point cloud registration (UPCR) method from the representation separation perspective. First, we model the input point cloud as a combination of pose-invariant representation and pose-related representation. Second, the pose-related representation is used to learn the relative pose wrt a "latent canonical shape" for the source and target point clouds respectively. Third, the rigid transformation is obtained from the above two learned relative poses. Our method not only filters out the disturbance in pose-invariant representation but also is robust to partial-to-partial point clouds or noise. Experiments on benchmark datasets demonstrate that our unsupervised method achieves comparable if not better performance than state-of-the-art supervised registration methods.
The task of semi-supervised video object segmentation (VOS) has been greatly advanced and state-of-the-art performance has been made by dense matching-based methods. The recent methods leverage space-time memory (STM) networks and learn to retrieve relevant information from all available sources, where the past frames with object masks form an external memory and the current frame as the query is segmented using the mask information in the memory. However, when forming the memory and performing matching, these methods only exploit the appearance information while ignoring the motion information. In this paper, we advocate the return of the \emph{motion information} and propose a motion uncertainty-aware framework (MUNet) for semi-supervised VOS. First, we propose an implicit method to learn the spatial correspondences between neighboring frames, building upon a correlation cost volume. To handle the challenging cases of occlusion and textureless regions during constructing dense correspondences, we incorporate the uncertainty in dense matching and achieve motion uncertainty-aware feature representation. Second, we introduce a motion-aware spatial attention module to effectively fuse the motion feature with the semantic feature. Comprehensive experiments on challenging benchmarks show that \textbf{\textit{using a small amount of data and combining it with powerful motion information can bring a significant performance boost}}. We achieve ${76.5\%}$ $\mathcal{J} \& \mathcal{F}$ only using DAVIS17 for training, which significantly outperforms the \textit{SOTA} methods under the low-data protocol. \textit{The code will be released.}