Existing datasets for manually labelled query-based video summarization are costly and thus small, limiting the performance of supervised deep video summarization models. Self-supervision can address the data sparsity challenge by using a pretext task and defining a method to acquire extra data with pseudo labels to pre-train a supervised deep model. In this work, we introduce segment-level pseudo labels from input videos to properly model both the relationship between a pretext task and a target task, and the implicit relationship between the pseudo label and the human-defined label. The pseudo labels are generated based on existing human-defined frame-level labels. To create more accurate query-dependent video summaries, a semantics booster is proposed to generate context-aware query representations. Furthermore, we propose mutual attention to help capture the interactive information between visual and textual modalities. Three commonly-used video summarization benchmarks are used to thoroughly validate the proposed approach. Experimental results show that the proposed video summarization algorithm achieves state-of-the-art performance.
相關內容
Skeleton-based action segmentation requires recognizing composable actions in untrimmed videos. Current approaches decouple this problem by first extracting local visual features from skeleton sequences and then processing them by a temporal model to classify frame-wise actions. However, their performances remain limited as the visual features cannot sufficiently express composable actions. In this context, we propose Latent Action Composition (LAC), a novel self-supervised framework aiming at learning from synthesized composable motions for skeleton-based action segmentation. LAC is composed of a novel generation module towards synthesizing new sequences. Specifically, we design a linear latent space in the generator to represent primitive motion. New composed motions can be synthesized by simply performing arithmetic operations on latent representations of multiple input skeleton sequences. LAC leverages such synthesized sequences, which have large diversity and complexity, for learning visual representations of skeletons in both sequence and frame spaces via contrastive learning. The resulting visual encoder has a high expressive power and can be effectively transferred onto action segmentation tasks by end-to-end fine-tuning without the need for additional temporal models. We conduct a study focusing on transfer-learning and we show that representations learned from pre-trained LAC outperform the state-of-the-art by a large margin on TSU, Charades, PKU-MMD datasets.
Streaming video clips with large-scale video tokens impede vision transformers (ViTs) for efficient recognition, especially in video action detection where sufficient spatiotemporal representations are required for precise actor identification. In this work, we propose an end-to-end framework for efficient video action detection (EVAD) based on vanilla ViTs. Our EVAD consists of two specialized designs for video action detection. First, we propose a spatiotemporal token dropout from a keyframe-centric perspective. In a video clip, we maintain all tokens from its keyframe, preserve tokens relevant to actor motions from other frames, and drop out the remaining tokens in this clip. Second, we refine scene context by leveraging remaining tokens for better recognizing actor identities. The region of interest (RoI) in our action detector is expanded into temporal domain. The captured spatiotemporal actor identity representations are refined via scene context in a decoder with the attention mechanism. These two designs make our EVAD efficient while maintaining accuracy, which is validated on three benchmark datasets (i.e., AVA, UCF101-24, JHMDB). Compared to the vanilla ViT backbone, our EVAD reduces the overall GFLOPs by 43% and improves real-time inference speed by 40% with no performance degradation. Moreover, even at similar computational costs, our EVAD can improve the performance by 1.1 mAP with higher resolution inputs. Code is available at //github.com/MCG-NJU/EVAD.
Different from traditional video cameras, event cameras capture asynchronous events stream in which each event encodes pixel location, trigger time, and the polarity of the brightness changes. In this paper, we introduce a novel graph-based framework for event cameras, namely SlideGCN. Unlike some recent graph-based methods that use groups of events as input, our approach can efficiently process data event-by-event, unlock the low latency nature of events data while still maintaining the graph's structure internally. For fast graph construction, we develop a radius search algorithm, which better exploits the partial regular structure of event cloud against k-d tree based generic methods. Experiments show that our method reduces the computational complexity up to 100 times with respect to current graph-based methods while keeping state-of-the-art performance on object recognition. Moreover, we verify the superiority of event-wise processing with our method. When the state becomes stable, we can give a prediction with high confidence, thus making an early recognition. Project page: \url{//zju3dv.github.io/slide_gcn/}.
The individual difference between subjects is significant in EEG-based emotion recognition, resulting in the difficulty of sharing the model across subjects. Previous studies use domain adaptation algorithms to minimize the global domain discrepancy while ignoring the class information, which may cause misalignment of subdomains and reduce model performance. This paper proposes a multi-subdomain adversarial network (MSAN) for cross-subject EEG-based emotion recognition. MSAN uses adversarial training to model the discrepancy in the global domain and subdomain to reduce the intra-class distance and enlarge the inter-class distance. In addition, MSAN initializes parameters through a pre-trained autoencoder to ensure the stability and convertibility of the model. The experimental results show that the accuracy of MSAN is improved by 30.02\% on the SEED dataset comparing with the nontransfer method.
In modern communication systems operating with Orthogonal Frequency-Division Multiplexing (OFDM), channel estimation requires minimal complexity with one-tap equalizers. However, this depends on cyclic prefixes, which must be sufficiently large to cover the channel impulse response. Conversely, the use of cyclic prefix (CP) decreases the useful information that can be conveyed in an OFDM frame, thereby degrading the spectral efficiency of the system. In this context, we study the impact of CPs on channel estimation with complex-valued neural networks (CVNNs). We show that the phase-transmittance radial basis function neural network offers superior results, in terms of required energy per bit, compared to classical minimum mean-squared error and least squares algorithms in scenarios without CP.
Sentiment analysis is the task of mining the authors' opinions about specific entities. It allows organizations to monitor different services in real time and act accordingly. Reputation is what is generally said or believed about people or things. Informally, reputation combines the measure of reliability derived from feedback, reviews, and ratings gathered from users, which reflect their quality of experience (QoE) and can either increase or harm the reputation of the provided services. In this study, we propose to perform sentiment analysis on web microservices reviews to exploit the provided information to assess and score the microservices' reputation. Our proposed approach uses the Long Short-Term Memory (LSTM) model to perform sentiment analysis and the Net Brand Reputation (NBR) algorithm to assess reputation scores for microservices. This approach is tested on a set of more than 10,000 reviews related to 15 Amazon Web microservices, and the experimental results have shown that our approach is more accurate than existing approaches, with an accuracy and precision of 93% obtained after applying an oversampling strategy and a resulting reputation score of the considered microservices community of 89%.
Moire patterns occur when capturing images or videos on screens, severely degrading the quality of the captured images or videos. Despite the recent progresses, existing video demoireing methods neglect the physical characteristics and formation process of moire patterns, significantly limiting the effectiveness of video recovery. This paper presents a unified framework, DTNet, a direction-aware and temporal-guided bilateral learning network for video demoireing. DTNet effectively incorporates the process of moire pattern removal, alignment, color correction, and detail refinement. Our proposed DTNet comprises two primary stages: Frame-level Direction-aware Demoireing and Alignment (FDDA) and Tone and Detail Refinement (TDR). In FDDA, we employ multiple directional DCT modes to perform the moire pattern removal process in the frequency domain, effectively detecting the prominent moire edges. Then, the coarse and fine-grained alignment is applied on the demoired features for facilitating the utilization of neighboring information. In TDR, we propose a temporal-guided bilateral learning pipeline to mitigate the degradation of color and details caused by the moire patterns while preserving the restored frequency information in FDDA. Guided by the aligned temporal features from FDDA, the affine transformations for the recovery of the ultimate clean frames are learned in TDR. Extensive experiments demonstrate that our video demoireing method outperforms state-of-the-art approaches by 2.3 dB in PSNR, and also delivers a superior visual experience.
Several recent works have directly extended the image masked autoencoder (MAE) with random masking into video domain, achieving promising results. However, unlike images, both spatial and temporal information are important for video understanding. This suggests that the random masking strategy that is inherited from the image MAE is less effective for video MAE. This motivates the design of a novel masking algorithm that can more efficiently make use of video saliency. Specifically, we propose a motion-guided masking algorithm (MGM) which leverages motion vectors to guide the position of each mask over time. Crucially, these motion-based correspondences can be directly obtained from information stored in the compressed format of the video, which makes our method efficient and scalable. On two challenging large-scale video benchmarks (Kinetics-400 and Something-Something V2), we equip video MAE with our MGM and achieve up to +$1.3\%$ improvement compared to previous state-of-the-art methods. Additionally, our MGM achieves equivalent performance to previous video MAE using up to $66\%$ fewer training epochs. Lastly, we show that MGM generalizes better to downstream transfer learning and domain adaptation tasks on the UCF101, HMDB51, and Diving48 datasets, achieving up to +$4.9\%$ improvement compared to baseline methods.
Currently, digital avatars can be created manually using human images as reference. Systems such as Bitmoji are excellent producers of detailed avatar designs, with hundreds of choices for customization. A supervised learning model could be trained to generate avatars automatically, but the hundreds of possible options create difficulty in securing non-noisy data to train a model. As a solution, we train a model to produce avatars from human images using tag-based annotations. This method provides better annotator agreement, leading to less noisy data and higher quality model predictions. Our contribution is an application of tag-based annotation to train a model for avatar face creation. We design tags for 3 different facial facial features offered by Bitmoji, and train a model using tag-based annotation to predict the nose.
Search in social networks such as Facebook poses different challenges than in classical web search: besides the query text, it is important to take into account the searcher's context to provide relevant results. Their social graph is an integral part of this context and is a unique aspect of Facebook search. While embedding-based retrieval (EBR) has been applied in eb search engines for years, Facebook search was still mainly based on a Boolean matching model. In this paper, we discuss the techniques for applying EBR to a Facebook Search system. We introduce the unified embedding framework developed to model semantic embeddings for personalized search, and the system to serve embedding-based retrieval in a typical search system based on an inverted index. We discuss various tricks and experiences on end-to-end optimization of the whole system, including ANN parameter tuning and full-stack optimization. Finally, we present our progress on two selected advanced topics about modeling. We evaluated EBR on verticals for Facebook Search with significant metrics gains observed in online A/B experiments. We believe this paper will provide useful insights and experiences to help people on developing embedding-based retrieval systems in search engines.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191