The objective of stylized speech-driven facial animation is to create animations that encapsulate specific emotional expressions. Existing methods often depend on pre-established emotional labels or facial expression templates, which may limit the necessary flexibility for accurately conveying user intent. In this research, we introduce a technique that enables the control of arbitrary styles by leveraging natural language as emotion prompts. This technique presents benefits in terms of both flexibility and user-friendliness. To realize this objective, we initially construct a Text-Expression Alignment Dataset (TEAD), wherein each facial expression is paired with several prompt-like descriptions. We propose an innovative automatic annotation method, supported by CahtGPT, to expedite the dataset construction, thereby eliminating the substantial expense of manual annotation. Following this, we utilize TEAD to train a CLIP-based model, termed ExpCLIP, which encodes text and facial expressions into semantically aligned style embeddings. The embeddings are subsequently integrated into the facial animation generator to yield expressive and controllable facial animations. Given the limited diversity of facial emotions in existing speech-driven facial animation training data, we further introduce an effective Expression Prompt Augmentation (EPA) mechanism to enable the animation generator to support unprecedented richness in style control. Comprehensive experiments illustrate that our method accomplishes expressive facial animation generation and offers enhanced flexibility in effectively conveying the desired style.
论文检索
输入标题、作者或关键词，从 2,866 篇学术成果中精准定位
Saccadic scanpath, a data representation of human visual behavior, has received broad interest in multiple domains. Scanpath is a complex eye-tracking data modality that includes the sequences of fixation positions and fixation duration, coupled with image information. However, previous methods usually face the spatial misalignment problem of fixation features and loss of critical temporal data (including temporal correlation and fixation duration). In this study, we propose a Transformer-based scanpath model, SpFormer, to alleviate these problems. First, we propose a fixation-centric paradigm to extract the aligned spatial fixation features and tokenize the scanpaths. Then, according to the visual working memory mechanism, we design a local meta attention to reduce the semantic redundancy of fixations and guide the model to focus on the meta scanpath. Finally, we progressively integrate the duration information and fuse it with the fixation features to solve the problem of ambiguous location with the Transformer block increasing. We conduct extensive experiments on four databases under three tasks. The SpFormer establishes new state-of-the-art results in distinct settings, verifying its flexibility and versatility in practical applications. The code can be obtained from https://github.com/wenqizhong/SpFormer.
Monocular 3D lane detection is essential for a reliable autonomous driving system and has recently been rapidly developing. Existing popular methods mainly employ a predefined 3D anchor for lane detection based on front-viewed (FV) space, aiming to mitigate the effects of view transformations. However, the perspective geometric distortion between FV and 3D space in this FV-based approach introduces extremely dense anchor designs, which ultimately leads to confusing lane representations. In this paper, we introduce a novel prior-guided perspective on lane detection and propose an end-to-end framework named PVALane, which utilizes 2D prior knowledge to achieve precise and efficient 3D lane detection. Since 2D lane predictions can provide strong priors for lane existence, PVALane exploits FV features to generate sparse prior anchors with potential lanes in 2D space. These dynamic prior anchors help PVALane to achieve distinct lane representations and effectively improve the precision of PVALane due to the reduced lane search space. Additionally, by leveraging these prior anchors and representing lanes in both FV and bird-eye-viewed (BEV) spaces, we effectively align and merge semantic and geometric information from FV and BEV features. Extensive experiments conducted on the OpenLane and ONCE-3DLanes datasets demonstrate the superior performance of our method compared to existing state-of-the-art approaches and exhibit excellent robustness.
Online contextual reasoning and association across consecutive video frames are critical to perceive instances in visual tracking. However, most current top-performing trackers persistently lean on sparse temporal relationships between reference and search frames via an offline mode. Consequently, they can only interact independently within each image-pair and establish limited temporal correlations. To alleviate the above problem, we propose a simple, flexible and effective video-level tracking pipeline, named ODTrack, which densely associates the contextual relationships of video frames in an online token propagation manner. ODTrack receives video frames of arbitrary length to capture the spatio-temporal trajectory relationships of an instance, and compresses the discrimination features (localization information) of a target into a token sequence to achieve frame-to-frame association. This new solution brings the following benefits: 1) the purified token sequences can serve as prompts for the inference in the next video frame, whereby past information is leveraged to guide future inference; 2) the complex online update strategies are effectively avoided by the iterative propagation of token sequences, and thus we can achieve more efficient model representation and computation. ODTrack achieves a new SOTA performance on seven benchmarks, while running at real-time speed. Code and models are available at https://github.com/GXNU-ZhongLab/ODTrack.
Graph Convolutional Networks (GCNs) and Transformers have been widely applied to skeleton-based human action recognition, with each offering unique advantages in capturing spatial relationships and long-range dependencies. However, for most GCN methods, the construction of topological structures relies solely on the spatial information of human joints, limiting their ability to directly capture richer spatio-temporal dependencies. Additionally, the self-attention modules of many Transformer methods lack topological structure information, restricting the robustness and generalization of the models. To address these issues, we propose a Joint Trajectory Graph (JTG) that integrates spatio-temporal information into a uniform graph structure. We also present a Joint Trajectory GraphFormer (JT-GraphFormer), which directly captures the spatio-temporal relationships among all joint trajectories for human action recognition. To better integrate topological information into spatio-temporal relationships, we introduce a Spatio-Temporal Dijkstra Attention (STDA) mechanism to calculate relationship scores for all the joints in JTG. Furthermore, we incorporate the Koopman operator into the classification stage to enhance the model's representation ability and classification performance. Experiments demonstrate that JT-GraphFormer achieves outstanding performance in human action recognition tasks, outperforming state-of-the-art methods on the NTU RGB+D, NTU RGB+D 120, and N-UCLA datasets.
Stable diffusion, a generative model used in text-to-image synthesis, frequently encounters resolution-induced composition problems when generating images of varying sizes. This issue primarily stems from the model being trained on pairs of single-scale images and their corresponding text descriptions. Moreover, direct training on images of unlimited sizes is unfeasible, as it would require an immense number of text-image pairs and entail substantial computational expenses. To overcome these challenges, we propose a two-stage pipeline named Any-Size-Diffusion (ASD), designed to efficiently generate well-composed HD images of any size, while minimizing the need for high-memory GPU resources. Specifically, the initial stage, dubbed Any Ratio Adaptability Diffusion (ARAD), leverages a selected set of images with a restricted range of ratios to optimize the text-conditional diffusion model, thereby improving its ability to adjust composition to accommodate diverse image sizes. To support the creation of images at any desired size, we further introduce a technique called Fast Seamless Tiled Diffusion (FSTD) at the subsequent stage. This method allows for the rapid enlargement of the ASD output to any high-resolution size, avoiding seaming artifacts or memory overloads. Experimental results on the LAION-COCO and MM-CelebA-HQ benchmarks demonstrate that ASD can produce well-structured images of arbitrary sizes, cutting down the inference time by 2X compared to the traditional tiled algorithm. The source code is available at https://github.com/ProAirVerse/Any-Size-Diffusion.
As a kind of 3D data, RGB-D images have been extensively used in object tracking, 3D reconstruction, remote sensing mapping, and other tasks. In the realm of computer vision, the significance of RGB-D images is progressively growing. However, the existing learning-based image compression methods usually process RGB images and depth images separately, which cannot entirely exploit the redundant information between the modalities, limiting the further improvement of the Rate-Distortion performance. With the goal of overcoming the defect, in this paper, we propose a learning-based dual-branch RGB-D image compression framework. Compared with traditional RGB domain compression scheme, a YUV domain compression scheme is presented for spatial redundancy removal. In addition, Intra-Modality Attention (IMA) and Cross-Modality Attention (CMA) are introduced for modal redundancy removal. For the sake of benefiting from cross-modal prior information, Context Prediction Module (CPM) and Context Fusion Module (CFM) are raised in the conditional entropy model which makes the context probability prediction more accurate. The experimental results demonstrate our method outperforms existing image compression methods in two RGB-D image datasets. Compared with BPG, our proposed framework can achieve up to 15% bit rate saving for RGB images.
Quad Bayer Joint Demosaicing and Denoising Based on Dual Encoder Network with Joint Residual Learning
PDF ↗The recent imaging technology Quad Bayer CFA brings better imaging PSNR and higher visual quality compared to traditional Bayer CFA, but also serious challenges for demosaicing and denoising during the ISP pipeline. In this paper, we propose a novel dual encoder network, namely DRNet, to achieve joint demosaicing and denoising for Quad Bayer CFA. The dual encoders are carefully designed in that one is mainly constructed by a joint residual block to jointly estimate the residuals for demosaicing and denoising separately. In contrast, the other one is started with a pixel modulation block which is specially designed to match the characteristics of Quad Bayer pattern for better feature extraction. We demonstrate the effectiveness of each proposed component through detailed ablation investigations. The comparison results on public benchmarks illustrate that our DRNet achieves an apparent performance gain~(0.38dB to the 2nd best) from the state-of-the-art method and balances performance and efficiency well. The experiments on real-world images show that the proposed method could enhance the reconstruction quality from the native ISP algorithm.
Obtaining large-scale radiology reports can be difficult for medical images due to ethical concerns, limiting the effectiveness of contrastive pre-training in the medical image domain and underscoring the need for alternative methods. In this paper, we propose eye-tracking as an alternative to text reports, as it allows for the passive collection of gaze signals without ethical issues. By tracking the gaze of radiologists as they read and diagnose medical images, we can understand their visual attention and clinical reasoning. When a radiologist has similar gazes for two medical images, it may indicate semantic similarity for diagnosis, and these images should be treated as positive pairs when pre-training a computer-assisted diagnosis (CAD) network through contrastive learning. Accordingly, we introduce the Medical contrastive Gaze Image Pre-training (McGIP) as a plug-and-play module for contrastive learning frameworks. McGIP uses radiologist gaze to guide contrastive pre-training. We evaluate our method using two representative types of medical images and two common types of gaze data. The experimental results demonstrate the practicality of McGIP, indicating its high potential for various clinical scenarios and applications.
Unifying Multi-Modal Uncertainty Modeling and Semantic Alignment for Text-to-Image Person Re-identification
PDF ↗Text-to-Image person re-identification (TI-ReID) aims to retrieve the images of target identity according to the given textual description. The existing methods in TI-ReID focus on aligning the visual and textual modalities through contrastive feature alignment or reconstructive masked language modeling (MLM). However, these methods parameterize the image/text instances as deterministic embeddings and do not explicitly consider the inherent uncertainty in pedestrian images and their textual descriptions, leading to limited image-text relationship expression and semantic alignment. To address the above problem, in this paper, we propose a novel method that unifies multi-modal uncertainty modeling and semantic alignment for TI-ReID. Specifically, we model the image and textual feature vectors of pedestrian as Gaussian distributions, where the multi-granularity uncertainty of the distribution is estimated by incorporating batch-level and identity-level feature variances for each modality. The multi-modal uncertainty modeling acts as a feature augmentation and provides richer image-text semantic relationship. Then we present a bi-directional cross-modal circle loss to more effectively align the probabilistic features between image and text in a self-paced manner. To further promote more comprehensive image-text semantic alignment, we design a task that complements the masked language modeling, focusing on the cross-modality semantic recovery of global masked token after cross-modal interaction. Extensive experiments conducted on three TI-ReID datasets highlight the effectiveness and superiority of our method over state-of-the-arts.
Image-level weakly supervised semantic segmentation has received increasing attention due to its low annotation cost. Existing methods mainly rely on Class Activation Mapping (CAM) to obtain pseudo-labels for training semantic segmentation models. In this work, we are the first to demonstrate that long-tailed distribution in training data can cause the CAM calculated through classifier weights over-activated for head classes and under-activated for tail classes due to the shared features among head- and tail- classes. This degrades pseudo-label quality and further influences final semantic segmentation performance. To address this issue, we propose a Shared Feature Calibration (SFC) method for CAM generation. Specifically, we leverage the class prototypes which carry positive shared features and propose a Multi-Scaled Distribution-Weighted (MSDW) consistency loss for narrowing the gap between the CAMs generated through classifier weights and class prototypes during training. The MSDW loss counterbalances over-activation and under-activation by calibrating the shared features in head-/tail-class classifier weights. Experimental results show that our SFC significantly improves CAM boundaries and achieves new state-of-the-art performances. The project is available at https://github.com/Barrett-python/SFC.
Knowledge distillation aims at reducing model size without compromising much performance. Recent work has applied it to large vision-language (VL) Transformers, and has shown that attention maps in the multi-head attention modules of vision-language Transformers contain extensive intra-modal and cross-modal co-reference relations to be distilled. The standard approach is to apply a one-to-one attention map distillation loss, i.e. the Teacher's first attention head instructs the Student's first head, the second teaches the second, and so forth, but this only works when the numbers of attention heads in the Teacher and Student are the same. To remove this constraint, we propose a new Attention Map Alignment Distillation (AMAD) method for Transformers with multi-head attention, which works for a Teacher and a Student with different numbers of attention heads. Specifically, we soft-align different heads in Teacher and Student attention maps using a cosine similarity weighting. The Teacher head contributes more to the Student heads for which it has a higher similarity weight. Each Teacher head contributes to all the Student heads by minimizing the divergence between the attention activation distributions for the soft-aligned heads. No head is left behind. This distillation approach operates like cross-attention. We experiment on distilling VL-T5 and BLIP, and apply AMAD loss on their T5, BERT, and ViT sub-modules. We show, under vision-language setting, that AMAD outperforms conventional distillation methods on VQA-2.0, COCO captioning, and Multi30K translation datasets. We further show that even without VL pre-training, the distilled VL-T5 models outperform corresponding VL pre-trained VL-T5 models that are further fine-tuned by ground-truth signals, and that fine-tuning distillation can also compensate to some degree for the absence of VL pre-training for BLIP models.
Human body orientation estimation (HBOE) aims to estimate the orientation of a human body relative to the camera’s frontal view. Despite recent advancements in this field, there still exist limitations in achieving fine-grained results. We identify certain defects and propose corresponding approaches as follows: 1). Existing datasets suffer from non-uniform angle distributions, resulting in sparse image data for certain angles. To provide comprehensive and high-quality data, we introduce RMOS (Rendered Model Orientation Set), a rendered dataset comprising 150K accurately labeled human instances with a wide range of orientations. 2). Directly using one-hot vector as labels may overlook the similarity between angle labels, leading to poor supervision. And converting the predictions from radians to degrees enlarges the regression error. To enhance supervision, we employ Laplace smoothing to vectorize the label, which contains more information. For fine-grained predictions, we adopt weighted Smooth-L1-loss to align predictions with the smoothed-label, thus providing robust supervision. 3). Previous works ignore body-part-specific information, resulting in coarse predictions. By employing local-window self-attention, our model could utilize different body part information for more precise orientation estimations. We validate the effectiveness of our method in the benchmarks with extensive experiments and show that our method outperforms state-of-the-art. Project is available at: https://github.com/Whalesong-zrs/Towards-Fine-grained-HBOE.
As an emerging neuromorphic camera with an asynchronous working mechanism, spike camera shows good potential for high-speed vision tasks. Each pixel in spike camera accumulates photons persistently and fires a spike whenever the accumulation exceeds a threshold. Such high-frequency fine-granularity photon recording facilitates the analysis and recovery of dynamic scenes with high-speed motion. This paper considers the optical flow estimation problem for spike cameras. Due to the Poisson nature of incoming photons, the occurrence of spikes is random and fluctuating, making conventional image matching inefficient. We propose a Hierarchical Spatial-Temporal (HiST) fusion module for spike representation to pursue reliable feature matching and develop a robust optical flow network, dubbed as HiST-SFlow. The HiST extracts features at multiple moments and hierarchically fuses the spatial-temporal information. We also propose an intra-moment filtering module to further extract the feature and suppress the influence of randomness in spikes. A scene loss is proposed to ensure that this hierarchical representation recovers the essential visual information in the scene. Experimental results demonstrate that the proposed method achieves state-of-the-art performance compared with the existing methods. The source codes are available at https://github.com/ruizhao26/HiST-SFlow.
Rethinking Two-Stage Referring Expression Comprehension: A Novel Grounding and Segmentation Method Modulated by Point
PDF ↗As a fundamental and challenging task in the vision and language domain, Referring Expression Comprehension (REC) has shown impressive improvements recently. However, for a complex task that couples the comprehension of abstract concepts and the localization of concrete instances, one-stage approaches are bottlenecked by computing and data resources. To obtain a low-cost solution, the prevailing two-stage approaches decouple REC into localization (region proposal) and comprehension (region-expression matching) at region-level, but the solution based on isolated regions cannot sufficiently utilize the context and is usually limited by the quality of proposals. Therefore, it is necessary to rebuild an efficient two-stage solution system. In this paper, we propose a point-based two-stage framework for REC, in which the two stages are redefined as point-based cross-modal comprehension and point-based instance localization. Specifically, we reconstruct the raw bounding box and segmentation mask into center and mass scores as soft ground-truth for measuring point-level cross-modal correlations. With the soft ground-truth, REC can be approximated as a binary classification problem, which fundamentally avoids the impact of isolated regions on the optimization process. Remarkably, the consistent metrics between center and mass scores allow our system to directly optimize grounding and segmentation by utilizing the same architecture. Experiments on multiple benchmarks show the feasibility and potential of our point-based paradigm. Our code available at https://github.com/VILAN-Lab/PBREC-MT.
Bio-inspired spike camera mimics the sampling principle of primate fovea. It presents high temporal resolution and dynamic range, showing great promise in fast-moving object recognition. However, the physical limit of CMOS technology in spike cameras still hinders their capability of recognizing ultra-high-speed moving objects, e.g., extremely fast motions cause blur during the imaging process of spike cameras. This paper presents the first theoretical analysis for the causes of spiking motion blur and proposes a robust representation that addresses this issue through temporal-spatial context learning. The proposed method leverages multi-span feature aggregation to capture temporal cues and employs residual deformable convolution to model spatial correlation among neighbouring pixels. Additionally, this paper contributes an original real-captured spiking recognition dataset consisting of 12,000 ultra-high-speed (equivalent speed > 500 km/h) moving objects. Experimental results show that the proposed method achieves 73.2% accuracy in recognizing 10 classes of ultra-high-speed moving objects, outperforming all existing spike-based recognition methods. Resources will be available at https://github.com/Evin-X/UHSR.
The completion of large occluded human body images poses a unique challenge for general image completion methods. The complex shape variations of human bodies make it difficult to establish a consistent understanding of their structures. Furthermore, as human vision is highly sensitive to human bodies, even slight artifacts can significantly compromise image fidelity. To address these challenges, we propose a large occluded human image completion (LOHC) model based on a novel image-prior cooperative completion strategy. Our model leverages human segmentation maps as a prior, and completes the image and prior simultaneously. Compared to the widely adopted prior-then-image completion strategy for object completion, this cooperative completion process fosters more effective interaction between the prior and image information. Our model consists of two stages. The first stage is a transformer-based auto-regressive network that predicts the overall structure of the missing area by generating a coarse completed image at a lower resolution. The second stage is a convolutional network that refines the coarse images. As the coarse result may not always be accurate, we propose a Dynamic Fusion Module (DFM) to selectively fuses the useful features from the coarse image with the original input at spatial and channel levels. Through extensive experiments, we demonstrate our method’s superior performance compared to state-of-the-art methods.
Multi-view camera-based 3D object detection has become popular due to its low cost, but accurately inferring 3D geometry solely from camera data remains challenging and may lead to inferior performance. Although distilling precise 3D geometry knowledge from LiDAR data could help tackle this challenge, the benefits of LiDAR information could be greatly hindered by the significant modality gap between different sensory modalities. To address this issue, we propose a Simulated multi-modal Distillation (SimDistill) method by carefully crafting the model architecture and distillation strategy. Specifically, we devise multi-modal architectures for both teacher and student models, including a LiDAR-camera fusion-based teacher and a simulated fusion-based student. Owing to the ``identical'' architecture design, the student can mimic the teacher to generate multi-modal features with merely multi-view images as input, where a geometry compensation module is introduced to bridge the modality gap. Furthermore, we propose a comprehensive multi-modal distillation scheme that supports intra-modal, cross-modal, and multi-modal fusion distillation simultaneously in the Bird's-eye-view space. Incorporating them together, our SimDistill can learn better feature representations for 3D object detection while maintaining a cost-effective camera-only deployment. Extensive experiments validate the effectiveness and superiority of SimDistill over state-of-the-art methods, achieving an improvement of 4.8% mAP and 4.1% NDS over the baseline detector. The source code will be released at https://github.com/ViTAE-Transformer/SimDistill.
Due to the ability to synthesize high-quality novel views, Neural Radiance Fields (NeRF) has been recently exploited to improve visual localization in a known environment. However, the existing methods mostly utilize NeRF for data augmentation to improve the regression model training, and their performances on novel viewpoints and appearances are still limited due to the lack of geometric constraints. In this paper, we propose a novel visual localization framework, i.e., PNeRFLoc, based on a unified point-based representation. On one hand, PNeRFLoc supports the initial pose estimation by matching 2D and 3D feature points as traditional structure-based methods; on the other hand, it also enables pose refinement with novel view synthesis using rendering-based optimization. Specifically, we propose a novel feature adaption module to close the gaps between the features for visual localization and neural rendering. To improve the efficacy and efficiency of neural rendering-based optimization, we also developed an efficient rendering-based framework with a warping loss function. Extensive experiments demonstrate that PNeRFLoc performs the best on the synthetic dataset when the 3D NeRF model can be well learned, and significantly outperforms all the NeRF-boosted localization methods with on-par SOTA performance on the real-world benchmark localization datasets. Project webpage: https://zju3dv.github.io/PNeRFLoc/.
Self-Distillation Regularized Connectionist Temporal Classification Loss for Text Recognition: A Simple Yet Effective Approach
PDF ↗Text recognition methods are gaining rapid development. Some advanced techniques, e.g., powerful modules, language models, and un- and semi-supervised learning schemes, consecutively push the performance on public benchmarks forward. However, the problem of how to better optimize a text recognition model from the perspective of loss functions is largely overlooked. CTC-based methods, widely used in practice due to their good balance between performance and inference speed, still grapple with accuracy degradation. This is because CTC loss emphasizes the optimization of the entire sequence target while neglecting to learn individual characters. We propose a self-distillation scheme for CTC-based model to address this issue. It incorporates a framewise regularization term in CTC loss to emphasize individual supervision, and leverages the maximizing-a-posteriori of latent alignment to solve the inconsistency problem that arises in distillation between CTC-based models. We refer to the regularized CTC loss as Distillation Connectionist Temporal Classification (DCTC) loss. DCTC loss is module-free, requiring no extra parameters, longer inference lag, or additional training data or phases. Extensive experiments on public benchmarks demonstrate that DCTC can boost text recognition model accuracy by up to 2.6%, without any of these drawbacks.