Video Object Segmentation Task Research Articles

LiDAR video object segmentation with dynamic kernel refinement

In this paper, we formalize memory- and tracking-based methods to perform the LiDAR-based Video Object Segmentation (VOS) task, which segments points of the specific 3D target (given in the first frame) in a LiDAR sequence. LiDAR-based VOS can directly provide target-aware geometric information for practical application scenarios like behavior analysis and anticipating danger. We first construct a LiDAR-based VOS dataset named KITTI-VOS based on SemanticKITTI, which acts as a testbed and facilitates comprehensive evaluations of algorithm performance. Next, we provide two types of baselines, i.e., memory-based and tracking-based baselines, to explore this task. Specifically, the first memory-based pipeline is built on a space–time memory network equipped with the non-local spatiotemporal attention-based memory bank. We further design a more potent variant to introduce the locality into the spatiotemporal attention module by local self-attention and cross-attention modules. For the second tracking-based baseline, we modify two representative 3D object tracking methods to adapt to LiDAR-based VOS tasks. Finally, we propose a refine module that takes mask priors and generates object-aware kernels, which could boost all the baselines’ performance. We evaluate the proposed methods on the dataset and demonstrate their effectiveness.

TokenCut: Segmenting Objects in Images and Videos with Self-supervised Transformer and Normalized Cut.

In this paper, we describe a graph-based algorithm that uses the features obtained by a self-supervised transformer to detect and segment salient objects in images and videos. With this approach, the image patches that compose an image or video are organised into a fully connected graph, in which the edge between each pair of patches is labeled with a similarity score based on the features learned by the transformer. Detection and segmentation of salient objects can then be formulated as a graph-cut problem and solved using the classical Normalized Cut algorithm. Despite the simplicity of this approach, it achieves state-of-the-art results on several common image and video detection and segmentation tasks. For unsupervised object discovery, this approach outperforms the competing approaches by a margin of 6.1%, 5.7%, and 2.6% when tested with the VOC07, VOC12, and COCO20K datasets. For the unsupervised saliency detection task in images, this method improves the score for Intersection over Union (IoU) by 4.4%, 5.6% and 5.2%. When tested with the ECSSD, DUTS, and DUT-OMRON datasets. This method also achieves competitive results for unsupervised video object segmentation tasks with the DAVIS, SegTV2, and FBMS datasets. Our implementation is available at https://www.m-psi.fr/Papers/TokenCut2022/.

Attention-guided Adversarial Attack for Video Object Segmentation

Video Object Segmentation (VOS) methods have made many breakthroughs with the help of the continuous development and advancement of deep learning. However, the deep learning model is vulnerable to malicious adversarial attacks, which mislead the model to make wrong decisions by adding adversarial perturbation that humans cannot perceive to the input image. Threats to deep learning models remind us that video object segmentation methods are also vulnerable to attacks, thereby threatening their security. Therefore, we study adversarial attacks on the VOS task to better identify the vulnerabilities of the VOS method, which in turn provides an opportunity to improve its robustness. In this paper, we propose an attention-guided adversarial attack method, which uses spatial attention blocks to capture features with global dependencies to construct correlations between consecutive video frames, and performs multipath aggregation to effectively integrate spatial-temporal perturbation, thereby guiding the deconvolution network to generate adversarial examples with strong attack capability. Specifically, the class loss function is designed to enable the deconvolution network to better activate noise in other regions and suppress the activation related to the object class based on the enhanced feature map of the object class. At the same time, attentional feature loss is designed to enhance the transferability against attack. The experimental results on the DAVIS dataset show that the proposed attention-guided adversarial attack method can significantly reduce the segmentation accuracy of OSVOS, and the J & F mean on DAVIS 2016 can reach 73.6% drop rate. The generated adversarial examples are also highly transferable to other video object segmentation models.

Efficient Long-Short Temporal Attention network for unsupervised Video Object Segmentation

Unsupervised Video Object Segmentation (VOS) aims at identifying the contours of primary foreground objects in videos without any prior knowledge. However, previous methods do not fully use spatial–temporal context and fail to tackle this challenging task in real-time. This motivates us to develop an efficient Long-Short Temporal Attention network (termed LSTA) for unsupervised VOS task from a holistic view. Specifically, LSTA consists of two dominant modules, i.e., Long Temporal Memory and Short Temporal Attention. The former captures the long-term global pixel relations of the past frames and the current frame, which models constantly present objects by encoding appearance pattern. Meanwhile, the latter reveals the short-term local pixel relations of one nearby frame and the current frame, which models moving objects by encoding motion pattern. To speedup the inference, the efficient projection and the locality-based sliding window are adopted to achieve nearly linear time complexity for the two light modules, respectively. Extensive empirical studies on several benchmarks have demonstrated promising performances of the proposed method with high efficiency.

Semi-supervised video object segmentation via an edge attention gated graph convolutional network

Video object segmentation (VOS) exhibits heavy occlusions, large deformation, and severe motion blur. While many remarkable convolutional neural networks are devoted to the VOS task, they often mis-identify background noise as the target or output coarse object boundaries, due to the failure of mining detail information and high-order correlations of pixels within the whole video. In this work, we propose an edge attention gated graph convolutional network (GCN) for VOS. The seed point initialization and graph construction stages construct a spatio-temporal graph of the video by exploring the spatial intra-frame correlation and the temporal inter-frame correlation of superpixels. The node classification stage identifies foreground superpixels by using an edge attention gated GCN which mines higher-order correlations between superpixels and propagates features among different nodes. The segmentation optimization stage optimizes the classification of foreground superpixels and reduces segmentation errors by using a global appearance model which captures the long-term stable feature of objects. In summary, the key contribution of our framework is twofold: (a) the spatio-temporal graph representation can propagate the seed points of the first frame to subsequent frames and facilitate our framework for the semi-supervised VOS task; (b) the edge attention gated GCN can learn the importance of each node with respect to both the neighboring nodes and the whole task with small number of layers. Experiments on Davis 2016 and Davis 2017 datasets show that our framework achieves the excellent performance with only small training samples (45 video sequences).

Video Object of Interest Segmentation

In this work, we present a new computer vision task named video object of interest segmentation (VOIS). Given a video and a target image of interest, our objective is to simultaneously segment and track all objects in the video that are relevant to the target image. This problem combines the traditional video object segmentation task with an additional image indicating the content that users are concerned with. Since no existing dataset is perfectly suitable for this new task, we specifically construct a large-scale dataset called LiveVideos, which contains 2418 pairs of target images and live videos with instance-level annotations. In addition, we propose a transformer-based method for this task. We revisit Swin Transformer and design a dual-path structure to fuse video and image features. Then, a transformer decoder is employed to generate object proposals for segmentation and tracking from the fused features. Extensive experiments on LiveVideos dataset show the superiority of our proposed method.

Motion cues guided feature aggregation and enhancement for video object segmentation

Video object segmentation (VOS) aims to separate unknown target objects from various given video sequences. Although many recent successful methods boosted the performance of VOS, especially those using deep convolution neural networks (CNNs), it is still difficult to aggregate deep features as well as motion cues effectively, which can be important to associate valid information of adjacent frames in video sequences. To tackle this problem, we propose a simple yet effective feature optimization method for VOS based on motion information. To achieve this, we construct a two-branch deep network and use computed motion cues (i.e., optical flow) to jointly optimize global and local interframe correlation information. Additionally, a clustering-based feature enhancement module is proposed to further fuse motion information and enhance the feature saliency of the target area. Optimized feature maps show a significant performance improvement in the final VOS tasks, especially those with rapid target movement. Experiments on the DAVIS16, DAVIS17, YouTube-Objects and YouTube-VOS datasets demonstrate that our simple feature aggregation and enhancement method for VOS improves segmentation accuracy effectively and gains an impressive result compared to many state-of-the-art methods.

An automatic quality evaluator for video object segmentation masks

Video object segmentation (VOS) has been a research hot-spot these years. However, evaluating the performance of different VOS methods requires labor-intensive and time-consuming manually labeled mask annotations, making it hard to validate the algorithm quality in field tests. In this paper, we tackle the problem of automatically measuring the mask quality for video object segmentation tasks without accessing manual annotations. We propose that with an elaborately designed network structure, we can extract quality-sensitive features to predict mask quality scores without ground-truth labels. To achieve this, we train an end-to-end convolutional neural network to capture the quality-sensitive features with both spatial reference and temporal reference. In the proposed Video Object Segmentation Evaluation Network, the VOSE-Net, the corresponding video frame and motion amplitude information are used for spatial and temporal references respectively. Instead of directly concatenating features for mask and references, we extract spatial quality cues with feature correlation, which is more rational and effective in this specific task. Taking in the segmented mask, its corresponding frame image and optical flow map, the VOSE-Net can provide an accurate quality estimation without the need for human intervention. To train and verify the proposed network, we construct a new dataset by using the DAVIS video segmentation benchmark and results from many public video object segmentation algorithms. We also demonstrate the robustness and usefulness of the proposed method on several applications, i.e. proposal selection, parameter optimization, arbitrary video mask evaluation. The experimental results and analysis show that the VOSE-Net is fast, effective and of practical use.

Learning multi-view visual correspondences with self-supervision

Stereo-based 3D reconstruction requires to match features across images captured from slightly different viewing angles to recover 3D coordinates of the image pixels. Despite the workload of collecting data, annotating matched pixels requires also heavy labor. As recent researches for self-supervised representation learning has gained great progress, learning multi-view visual correspondences from large scale raw videos serves as an alternative. However, existing methods which benefit from contrastive learning tend to neglect false negative samples when matching between adjacent frames in a video, leading to sub-optimal optimization for visual features. In this paper, we propose a contrastive learning framework that construct self-supervision by semi-global visual correspondence to alleviate learning degradation when false negatives are involved in training. Our learning framework consists of pixel-level contrastive learning via patch reconstruction and patch-level contrastive learning cross videos. We also introduce saliency guidance to extract salient regions from video frames to further reduce potential false negatives. By optimizing the model with the proposed semi-global contrastive learning method, learned representations are forced to be discriminative and robust. Experiments demonstrate that our proposed method outperforms previous self-supervised methods on video object segmentation tasks. Moreover, when compared to fully-supervised algorithms designed for specific tasks, our proposed method also achieves competitive results.

Physarum Powered Differentiable Linear Programming Layers and Applications

Consider a learning algorithm, which involves an internal call to an optimization routine such as a generalized eigenvalue problem, a cone programming problem or even sorting. Integrating such a method as layers within a trainable deep network in a numerically stable way is not simple – for instance, only recently, strategies have emerged for eigendecomposition and differentiable sorting. We propose an efficient and differentiable solver for general linear programming problems which can be used in a plug and play manner within deep neural networks as a layer. Our development is inspired by a fascinating but not widely used link between dynamics of slime mold (physarum) and mathematical optimization schemes such as steepest descent. We describe our development and demonstrate the use of our solver in a video object segmentation task and meta-learning for few-shot learning. We review the relevant known results and provide a technical analysis describing its applicability for our use cases. Our solver performs comparably with a customized projected gradient descent method on the first task and outperforms the very recently proposed differentiable CVXPY solver on the second task. Experiments show that our solver converges quickly without the need for a feasible initial point. Interestingly, our scheme is easy to implement and can easily serve as layers whenever a learning procedure needs a fast approximate solution to a LP, within a larger network.

Meta-VOS: Learning to Adapt Online Target-Specific Segmentation.

The task of video object segmentation is a fundamental but challenging problem in the field of computer vision. To deal with large variations in target objects and background clutter, we propose an online adaptive video object segmentation (VOS) framework, named Meta-VOS, that learns to adapt the target-specific segmentation. Meta-VOS builds an online adaptive learning process by exploiting cumulative expertise after searching for confidence patterns across different videos/frames, and then dynamically improves the model learning from two aspects: Meta-seg learner (i.e., module updating) and Meta-seg criterion (i.e., rule of expertise). As our goal is to rapidly determine which patterns best represent the essential characteristics of specific targets in a video, Meta-seg learner is introduced to adaptively learn to update the parameters and hyperparameters of segmentation network in very few gradient descent steps. Furthermore, a Meta-seg criterion of learned expertise, which is constructed to evaluate the Meta-seg learner for the online adaptation of the segmentation network, can confidently online update positive/negative patterns under the guidance of motion cues, object appearances and learned knowledge. Comprehensive evaluations on several benchmark datasets demonstrate the superiority of our proposed Meta-VOS when compared with other state-of-the-art methods applied to the VOS problem.

Space-Time Memory Networks for Video Object Segmentation With User Guidance.

We propose a novel and unified solution for user-guided video object segmentation tasks. In this work, we consider two scenarios of user-guided segmentation: semi-supervised and interactive segmentation. Due to the nature of the problem, available cues - video frame(s) with object masks (or scribbles) - become richer with the intermediate predictions (or additional user inputs). However, the existing methods make it impossible to fully exploit this rich source of information. We resolve the issue by leveraging memory networks and learning to read relevant information from all available sources. In the semi-supervised scenario, the previous frames with object masks form an external memory, and the current frame as the query is segmented using the information in the memory. Similarly, to work with user interactions, the frames that are given user inputs form the memory that guides segmentation. Internally, the query and the memory are densely matched in the feature space, covering all the space-time pixel locations in a feed-forward fashion. The abundant use of the guidance information allows us to better handle challenges such as appearance changes and occlusions. We validate our method on the latest benchmark sets and achieve state-of-the-art performance along with a fast runtime.

Zero-Shot Video Object Segmentation With Co-Attention Siamese Networks.

We introduce a novel network, called CO-attention siamese network (COSNet), to address the zero-shot video object segmentation task in a holistic fashion. We exploit the inherent correlation among video frames and incorporate a global co-attention mechanism to further improve the state-of-the-art deep learning based solutions that primarily focus on learning discriminative foreground representations over appearance and motion in short-term temporal segments. The co-attention layers in COSNet provide efficient and competent stages for capturing global correlations and scene context by jointly computing and appending co-attention responses into a joint feature space. COSNet is a unified and end-to-end trainable framework where different co-attention variants can be derived for capturing diverse properties of the learned joint feature space. We train COSNet with pairs (or groups) of video frames, and this naturally augments training data and allows increased learning capacity. During the segmentation stage, the co-attention model encodes useful information by processing multiple reference frames together, which is leveraged to infer the frequently reappearing and salient foreground objects better. Our extensive experiments over three large benchmarks demonstrate that COSNet outperforms the current alternatives by a large margin. Our implementations are available at https://github.com/carrierlxk/COSNet.

Divided attention

Researchers from Nanjing University of Information Science and Technology (NUIST) present an attention-modulating network for video object segmentation with an advanced attention modulator to efficiently modulate a segmentation model to focus on a specific object of interest. The group employ a focal loss that distinguishes simple samples from more difficult ones to accelerate the convergence of network training to achieve state-of-the-art segmentation performance. Video object segmentation (VOS) is a fundamental task in computer vision, with important applications in video editing, robotics, and self-driving cars. VOS tasks are mainly categorised into unsupervised and semi-supervised classifications. The former seeks to find and segment the salient targets in the videos completely without supervision, with the algorithm itself deciding what the main segmentation is. The latter aims at segmenting an object instance throughout the entire video sequence given only the object mask on the first frame. This can be observed as a pixel-level object tracking problem. Semi-supervised VOS can be subdivided into single-object segmentation and multi-object segmentation. In the team's Letter, they focus on semi-supervised VOS. Deep learning for VOS has gained attention in the research community in recent years. Existing semi-supervised VOS techniques work by constructing deep networks and fine-tuning a pre-trained classifier on a given ground truth in the first frame during online testing. This online fine-tuning of a classifier during testing has been shown to significantly improve accuracy. Illustrative diagram of the proposed segmentation model and approach. Segmentation results. The team conduct an attention-modulating network for the semi-supervised VOS task. Co-author Kaihua Zhang elaborates on the process: “We designed an efficient visual and spatial attention modulator based on the semantic information of the annotated object in the first frame and the spatial information of predicted object mask in the previous frame, respectively, to fast module the segmentation model to focus on the specific object of interest. Then we design a SCAM architecture which includes a channel attention module and a spatial attention module and inject it into segmentation model to further refine its feature maps. In addition, we construct a feature pyramid attention module to mine context information of different scales to solve the problem of multi-scale segmentation. Most existing methods rely on fine-tuning models using first-frame annotations and are time-consuming, making them unsuitable for most practical applications. To address this issue, the proposed approach developed an attention-modulating network to focus on the appearance of a specific object instance in one single feed-forward pass without fine-tuning. Compared with other methods, this method has achieved state-of-art performance on the DAVIS2017 dataset by using attention-modulators, feature attention pyramid modules and focal loss. In order to overcome a sample imbalance problem, reference was made to focal loss which can accelerate the convergence of network training, thus helping to distinguish between difficult and simple samples. VOS remains challenging due to occlusions, fast motion, deformation, and significant appearance variations over time. This method conducts a visual attention modulator to extract semantic information such as category, color and shape from the first frame. The spatial attention modulator fits the predicted location of object masks in the previous frame as a spatial prior to guide the segmentation network to focus on the regions where that target is most likely to appear in the current frame. To solve the multi-scales of segmentation objects, feature pyramid attention modules mined the context information of different scales, achieving better pixel-level attention for the high-level feature maps. The proposed VOS approach is fast, which facilitates many applications, such as interactive video editing and augmented reality. It may be applied to video understanding models in the short term, and after long-term development, it may be applied to robotics, and self-driving cars. Kaihua Zhang notes on his groups future work: “Experiments show that our algorithm performs erroneous instance segmentation when faced with the challenge of occluding each other between similar objects. To tackle this problem, we will leverage a position-sensitive embedding which is capable of distinguishing the pixels of similar objects. We have also found that solving VOS with multiple instances requires template matching to deal with occlusion and temporal propagation to ensure temporal continuity; otherwise the segmentation instance would be lost. Thus, we will use the re-identification module to retrieve lost instances and take its frame as the starting point and use the mask propagation module to bi-directionally recover the lost instances.” The development of VOS in the next decade will achieve higher precision while meeting real-time application requirements. At present, the cost of manual annotation of pixel-level VOS data sets is too expensive, so cheaper large-scale VOS data sets are expected in the future.

Lucid Data Dreaming for Video Object Segmentation

Convolutional networks reach top quality in pixel-level video object segmentation but require a large amount of training data (1k–100k) to deliver such results. We propose a new training strategy which achieves state-of-the-art results across three evaluation datasets while using 20,times –1000,times less annotated data than competing methods. Our approach is suitable for both single and multiple object segmentation. Instead of using large training sets hoping to generalize across domains, we generate in-domain training data using the provided annotation on the first frame of each video to synthesize—“lucid dream” (in a lucid dream the sleeper is aware that he or she is dreaming and is sometimes able to control the course of the dream)—plausible future video frames. In-domain per-video training data allows us to train high quality appearance- and motion-based models, as well as tune the post-processing stage. This approach allows to reach competitive results even when training from only a single annotated frame, without ImageNet pre-training. Our results indicate that using a larger training set is not automatically better, and that for the video object segmentation task a smaller training set that is closer to the target domain is more effective. This changes the mindset regarding how many training samples and general “objectness” knowledge are required for the video object segmentation task.

Online Meta Adaptation for Fast Video Object Segmentation

Conventional deep neural networks based video object segmentation (VOS) methods are dominated by heavily fine-tuning a segmentation model on the first frame of a given video, which is time-consuming and inefficient. In this paper, we propose a novel method which rapidly adapts a base segmentation model to new video sequences with only a couple of model-update iterations, without sacrificing performance. Such attractive efficiency benefits from the meta-learning paradigm which leads to a meta-segmentation model and a novel continuous learning approach which enables online adaptation of the segmentation model. Concretely, we train a meta-learner on multiple VOS tasks such that the meta model can capture their common knowledge and gains the ability to fast adapt the segmentation model to new video sequences. Furthermore, to deal with unique challenges of VOS tasks from temporal variations in the video, e.g., object motion and appearance changes, we propose a principled online adaptation approach that continuously adapts the segmentation model across video frames by exploiting temporal context effectively, providing robustness to annoying temporal variations. Integrating the meta-learner with the online adaptation approach, the proposed VOS model achieves competitive performance against the state-of-the-arts and moreover provides faster per-frame processing speed.

Joint Video Object Discovery and Segmentation by Coupled Dynamic Markov Networks.

It is a challenging task to extract segmentation mask of a target from a single noisy video, which involves object discovery coupled with segmentation. To solve this challenge, we present a method to jointly discover and segment an object from a noisy video, where the target disappears intermittently throughout the video. Previous methods either only fulfill video object discovery, or video object segmentation presuming the existence of the object in each frame. We argue that jointly conducting the two tasks in a unified way will be beneficial. In other words, video object discovery and video object segmentation tasks can facilitate each other. To validate this hypothesis, we propose a principled probabilistic model, where two dynamic Markov networks are coupled-one for discovery and the other for segmentation. When conducting the Bayesian inference on this model using belief propagation, the bi-directional message passing reveals a clear collaboration between these two inference tasks. We validated our proposed method in five data sets. The first three video data sets, i.e., the SegTrack data set, the YouTube-objects data set, and the Davis data set, are not noisy, where all video frames contain the objects. The two noisy data sets, i.e., the XJTU-Stevens data set, and the Noisy-ViDiSeg data set, newly introduced in this paper, both have many frames that do not contain the objects. When compared with state of the art, it is shown that although our method produces inferior results on video data sets without noisy frames, we are able to obtain better results on video data sets with noisy frames.

Unsupervised video object segmentation using conditional random fields

In this work, we propose a graph-based superpixel segmentation technique to perform spatiotemporal oversegmentation of videos. The generated superpixels are post-processed by applying a straightforward threshold-based foreground separation model. These superpixels are used in a conditional random field, and a potential function is defined, which is solved using energy minimization techniques to produce a final segmentation. Experiments on two datasets containing over 24 videos demonstrate that our method produces competitive or better results for the video object segmentation task over the state-of-the-art algorithms.

Video Salient Object Detection Using Spatiotemporal Deep Features.

This paper presents a method for detecting salient objects in videos, where temporal information in addition to spatial information is fully taken into account. Following recent reports on the advantage of deep features over conventional handcrafted features, we propose a new set of spatiotemporal deep (STD) features that utilize local and global contexts over frames. We also propose new spatiotemporal conditional random field (STCRF) to compute saliency from STD features. STCRF is our extension of CRF to the temporal domain and describes the relationships among neighboring regions both in a frame and over frames. STCRF leads to temporally consistent saliency maps over frames, contributing to accurate detection of salient objects' boundaries and noise reduction during detection. Our proposed method first segments an input video into multiple scales and then computes a saliency map at each scale level using STD features with STCRF. The final saliency map is computed by fusing saliency maps at different scale levels. Our experiments, using publicly available benchmark datasets, confirm that the proposed method significantly outperforms the state-of-the-art methods. We also applied our saliency computation to the video object segmentation task, showing that our method outperforms existing video object segmentation methods.