Min Wang

Zhiyang Guo, Wengang Zhou, Min Wang et al.

We propose a novel framework to reconstruct accurate appearance and geometry with neural radiance fields (NeRF) for interacting hands, enabling the rendering of photo-realistic images and videos for gesture animation from arbitrary views. Given multi-view images of a single hand or interacting hands, an off-the-shelf skeleton estimator is first employed to parameterize the hand poses. Then we design a pose-driven deformation field to establish correspondence from those different poses to a shared canonical space, where a pose-disentangled NeRF for one hand is optimized. Such unified modeling efficiently complements the geometry and texture cues in rarely-observed areas for both hands. Meanwhile, we further leverage the pose priors to generate pseudo depth maps as guidance for occlusion-aware density learning. Moreover, a neural feature distillation method is proposed to achieve cross-domain alignment for color optimization. We conduct extensive experiments to verify the merits of our proposed HandNeRF and report a series of state-of-the-art results both qualitatively and quantitatively on the large-scale InterHand2.6M dataset.

Ziyin Zhang, Ning Lu, Minghui Liao et al.

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.

Xiaohan Lei, Min Wang, Wengang Zhou et al.

In embodied vision, Instance ImageGoal Navigation (IIN) requires an agent to locate a specific object depicted in a goal image within an unexplored environment. The primary challenge of IIN arises from the need to recognize the target object across varying viewpoints while ignoring potential distractors. Existing map-based navigation methods typically use Bird's Eye View (BEV) maps, which lack detailed texture representation of a scene. Consequently, while BEV maps are effective for semantic-level visual navigation, they are struggling for instance-level tasks. To this end, we propose a new framework for IIN, Gaussian Splatting for Visual Navigation (GaussNav), which constructs a novel map representation based on 3D Gaussian Splatting (3DGS). The GaussNav framework enables the agent to memorize both the geometry and semantic information of the scene, as well as retain the textural features of objects. By matching renderings of similar objects with the target, the agent can accurately identify, ground, and navigate to the specified object. Our GaussNav framework demonstrates a significant performance improvement, with Success weighted by Path Length (SPL) increasing from 0.347 to 0.578 on the challenging Habitat-Matterport 3D (HM3D) dataset. The source code is publicly available at the link: https://github.com/XiaohanLei/GaussNav.

Ting-Wei Zhou, Xi-Le Zhao, Jian-Li Wang et al.

Recently, the transform-based tensor representation has attracted increasing attention in multimedia data (e.g., images and videos) recovery problems, which consists of two indispensable components, i.e., transform and characterization. Previously, the development of transform-based tensor representation mainly focuses on the transform aspect. Although several attempts consider using shallow matrix factorization (e.g., singular value decomposition and negative matrix factorization) to characterize the frontal slices of transformed tensor (termed as latent tensor), the faithful characterization aspect is underexplored. To address this issue, we propose a unified Deep Tensor Representation (termed as DTR) framework by synergistically combining the deep latent generative module and the deep transform module. Especially, the deep latent generative module can faithfully generate the latent tensor as compared with shallow matrix factorization. The new DTR framework not only allows us to better understand the classic shallow representations, but also leads us to explore new representation. To examine the representation ability of the proposed DTR, we consider the representative multi-dimensional data recovery task and suggest an unsupervised DTR-based multi-dimensional data recovery model. Extensive experiments demonstrate that DTR achieves superior performance compared to state-of-the-art methods in both quantitative and qualitative aspects, especially for fine details recovery.

Yingda Guan, Zhengyang Feng, Huiying Chang et al.

We present SDTracker, a method that harnesses the potential of synthetic data for multi-object tracking of real-world scenes in a domain generalization and semi-supervised fashion. First, we use the ImageNet dataset as an auxiliary to randomize the style of synthetic data. With out-of-domain data, we further enforce pyramid consistency loss across different "stylized" images from the same sample to learn domain invariant features. Second, we adopt the pseudo-labeling method to effectively utilize the unlabeled MOT17 training data. To obtain high-quality pseudo-labels, we apply proximal policy optimization (PPO2) algorithm to search confidence thresholds for each sequence. When using the unlabeled MOT17 training set, combined with the pure-motion tracking strategy upgraded via developed post-processing, we finally reach 61.4 HOTA.

Mingxuan Gao, Min Wang, Maoyin Chen

Deep learning has shown the great power in the field of fault detection. However, for incipient faults with tiny amplitude, the detection performance of the current deep learning networks (DLNs) is not satisfactory. Even if prior information about the faults is utilized, DLNs can't successfully detect faults 3, 9 and 15 in Tennessee Eastman process (TEP). These faults are notoriously difficult to detect, lacking effective detection technologies in the field of fault detection. In this work, we propose Autoencoder-assisted Feature Ensemble Net (AE-FENet): a deep feature ensemble framework that uses the unsupervised autoencoder to conduct the feature transformation. Compared with the principle component analysis (PCA) technique adopted in the original Feature Ensemble Net (FENet), autoencoder can mine more exact features on incipient faults, which results in the better detection performance of AE-FENet. With same kinds of basic detectors, AE-FENet achieves a state-of-the-art average accuracy over 96% on faults 3, 9 and 15 in TEP, which represents a significant enhancement in performance compared to other methods. Plenty of experiments have been done to extend our framework, proving that DLNs can be utilized efficiently within this architecture.

Hao Yang, Min Wang, Zhengfei Yu et al.

It is well-known that deep neural networks (DNNs) have shown remarkable success in many fields. However, when adding an imperceptible magnitude perturbation on the model input, the model performance might get rapid decrease. To address this issue, a randomness technique has been proposed recently, named Stochastic Neural Networks (SNNs). Specifically, SNNs inject randomness into the model to defend against unseen attacks and improve the adversarial robustness. However, existed studies on SNNs mainly focus on injecting fixed or learnable noises to model weights/activations. In this paper, we find that the existed SNNs performances are largely bottlenecked by the feature representation ability. Surprisingly, simply maximizing the variance per dimension of the feature distribution leads to a considerable boost beyond all previous methods, which we named maximize feature distribution variance stochastic neural network (MFDV-SNN). Extensive experiments on well-known white- and black-box attacks show that MFDV-SNN achieves a significant improvement over existing methods, which indicates that it is a simple but effective method to improve model robustness.

Sibo Qiao, Shanchen Pang, Gang Luo et al.

Echocardiography is a powerful prenatal examination tool for early diagnosis of fetal congenital heart diseases (CHDs). The four-chamber (FC) view is a crucial and easily accessible ultrasound (US) image among echocardiography images. Automatic analysis of FC views contributes significantly to the early diagnosis of CHDs. The first step to automatically analyze fetal FC views is locating the fetal four crucial chambers of heart in a US image. However, it is a greatly challenging task due to several key factors, such as numerous speckles in US images, the fetal cardiac chambers with small size and unfixed positions, and category indistinction caused by the similarity of cardiac chambers. These factors hinder the process of capturing robust and discriminative features, hence destroying fetal cardiac anatomical chambers precise localization. Therefore, we first propose a multistage residual hybrid attention module (MRHAM) to improve the feature learning. Then, we present an improved YOLOv4 detection model, namely MRHAM-YOLOv4-Slim. Specially, the residual identity mapping is replaced with the MRHAM in the backbone of MRHAM-YOLOv4-Slim, accurately locating the four important chambers in fetal FC views. Extensive experiments demonstrate that our proposed method outperforms current state-of-the-art, including the precision of 0.919, the recall of 0.971, the F1 score of 0.944, the mAP of 0.953, and the frames per second (FPS) of 43.

Kai-Xin Gao, Xiao-Lei Liu, Zheng-Hai Huang et al.

Second-order optimization methods have the ability to accelerate convergence by modifying the gradient through the curvature matrix. There have been many attempts to use second-order optimization methods for training deep neural networks. Inspired by diagonal approximations and factored approximations such as Kronecker-Factored Approximate Curvature (KFAC), we propose a new approximation to the Fisher information matrix (FIM) called Trace-restricted Kronecker-factored Approximate Curvature (TKFAC) in this work, which can hold the certain trace relationship between the exact and the approximate FIM. In TKFAC, we decompose each block of the approximate FIM as a Kronecker product of two smaller matrices and scaled by a coefficient related to trace. We theoretically analyze TKFAC's approximation error and give an upper bound of it. We also propose a new damping technique for TKFAC on convolutional neural networks to maintain the superiority of second-order optimization methods during training. Experiments show that our method has better performance compared with several state-of-the-art algorithms on some deep network architectures.