UniSA STEM, University of South Australia, Adelaide, SA, Australia
Abstract:Thermal infrared tracking is an essential topic in computer vision tasks because of its advantage of all-weather imaging. However, most conventional methods utilize only hand-crafted features, while deep learning-based correlation filtering methods are limited by simple correlation operations. Transformer-based methods ignore temporal and coordinate information, which is critical for TIR tracking that lacks texture and color information. In this paper, to address these issues, we apply natural language modeling to TIR tracking and propose a novel model called NLMTrack, which enhances the utilization of coordinate and temporal information. NLMTrack applies an encoder that unifies feature extraction and feature fusion, which simplifies the TIR tracking pipeline. To address the challenge of low detail and low contrast in TIR images, on the one hand, we design a multi-level progressive fusion module that enhances the semantic representation and incorporates multi-scale features. On the other hand, the decoder combines the TIR features and the coordinate sequence features using a causal transformer to generate the target sequence step by step. Moreover, we explore an adaptive loss aimed at elevating tracking accuracy and a simple template update strategy to accommodate the target's appearance variations. Experiments show that NLMTrack achieves state-of-the-art performance on multiple benchmarks. The Code is publicly available at \url{https://github.com/ELOESZHANG/NLMTrack}.
Abstract:Facial nerve segmentation is crucial for preoperative path planning in cochlear implantation surgery. Recently, researchers have proposed some segmentation methods, such as atlas-based and deep learning-based methods. However, since the facial nerve is a tubular organ with a diameter of only 1.0-1.5mm, it is challenging to locate and segment the facial nerve in CT scans. In this work, we propose an uncertainty-aware dualstream network (UADSN). UADSN consists of a 2D segmentation stream and a 3D segmentation stream. Predictions from two streams are used to identify uncertain regions, and a consistency loss is employed to supervise the segmentation of these regions. In addition, we introduce channel squeeze & spatial excitation modules into the skip connections of U-shaped networks to extract meaningful spatial information. In order to consider topologypreservation, a clDice loss is introduced into the supervised loss function. Experimental results on the facial nerve dataset demonstrate the effectiveness of UADSN and our submodules.
Abstract:Semi-supervised learning (SSL) has shown considerable potential in medical image segmentation, primarily leveraging consistency regularization and pseudo-labeling. However, many SSL approaches only pay attention to low-level consistency and overlook the significance of pseudo-label reliability. Therefore, in this work, we propose an adversarial self-training consistency framework (AstMatch). Firstly, we design an adversarial consistency regularization (ACR) approach to enhance knowledge transfer and strengthen prediction consistency under varying perturbation intensities. Second, we apply a feature matching loss for adversarial training to incorporate high-level consistency regularization. Additionally, we present the pyramid channel attention (PCA) and efficient channel and spatial attention (ECSA) modules to improve the discriminator's performance. Finally, we propose an adaptive self-training (AST) approach to ensure the pseudo-labels' quality. The proposed AstMatch has been extensively evaluated with cutting-edge SSL methods on three public-available datasets. The experimental results under different labeled ratios indicate that AstMatch outperforms other existing methods, achieving new state-of-the-art performance. Our code will be available at https://github.com/GuanghaoZhu663/AstMatch.
Abstract:Micro-expressions (MEs) are subtle facial movements that occur spontaneously when people try to conceal the real emotions. Micro-expression recognition (MER) is crucial in many fields, including criminal analysis and psychotherapy. However, MER is challenging since MEs have low intensity and ME datasets are small in size. To this end, a three-stream temporal-shift attention network based on self-knowledge distillation (SKD-TSTSAN) is proposed in this paper. Firstly, to address the low intensity of ME muscle movements, we utilize learning-based motion magnification modules to enhance the intensity of ME muscle movements. Secondly, we employ efficient channel attention (ECA) modules in the local-spatial stream to make the network focus on facial regions that are highly relevant to MEs. In addition, temporal shift modules (TSMs) are used in the dynamic-temporal stream, which enables temporal modeling with no additional parameters by mixing ME motion information from two different temporal domains. Furthermore, we introduce self-knowledge distillation (SKD) into the MER task by introducing auxiliary classifiers and using the deepest section of the network for supervision, encouraging all blocks to fully explore the features of the training set. Finally, extensive experiments are conducted on four ME datasets: CASME II, SAMM, MMEW, and CAS(ME)3. The experimental results demonstrate that our SKD-TSTSAN outperforms other existing methods and achieves new state-of-the-art performance. Our code will be available at https://github.com/GuanghaoZhu663/SKD-TSTSAN.
Abstract:Video generation is a challenging yet pivotal task in various industries, such as gaming, e-commerce, and advertising. One significant unresolved aspect within T2V is the effective visualization of text within generated videos. Despite the progress achieved in Text-to-Video~(T2V) generation, current methods still cannot effectively visualize texts in videos directly, as they mainly focus on summarizing semantic scene information, understanding, and depicting actions. While recent advances in image-level visual text generation show promise, transitioning these techniques into the video domain faces problems, notably in preserving textual fidelity and motion coherence. In this paper, we propose an innovative approach termed Text-Animator for visual text video generation. Text-Animator contains a text embedding injection module to precisely depict the structures of visual text in generated videos. Besides, we develop a camera control module and a text refinement module to improve the stability of generated visual text by controlling the camera movement as well as the motion of visualized text. Quantitative and qualitative experimental results demonstrate the superiority of our approach to the accuracy of generated visual text over state-of-the-art video generation methods. The project page can be found at https://laulampaul.github.io/text-animator.html.
Abstract:Foundation models (FMs) are revolutionizing the analysis and understanding of remote sensing (RS) scenes, including aerial RGB, multispectral, and SAR images. However, hyperspectral images (HSIs), which are rich in spectral information, have not seen much application of FMs, with existing methods often restricted to specific tasks and lacking generality. To fill this gap, we introduce HyperSIGMA, a vision transformer-based foundation model for HSI interpretation, scalable to over a billion parameters. To tackle the spectral and spatial redundancy challenges in HSIs, we introduce a novel sparse sampling attention (SSA) mechanism, which effectively promotes the learning of diverse contextual features and serves as the basic block of HyperSIGMA. HyperSIGMA integrates spatial and spectral features using a specially designed spectral enhancement module. In addition, we construct a large-scale hyperspectral dataset, HyperGlobal-450K, for pre-training, which contains about 450K hyperspectral images, significantly surpassing existing datasets in scale. Extensive experiments on various high-level and low-level HSI tasks demonstrate HyperSIGMA's versatility and superior representational capability compared to current state-of-the-art methods. Moreover, HyperSIGMA shows significant advantages in scalability, robustness, cross-modal transferring capability, and real-world applicability.
Abstract:LiDAR-based sparse 3D object detection plays a crucial role in autonomous driving applications due to its computational efficiency advantages. Existing methods either use the features of a single central voxel as an object proxy, or treat an aggregated cluster of foreground points as an object proxy. However, the former lacks the ability to aggregate contextual information, resulting in insufficient information expression in object proxies. The latter relies on multi-stage pipelines and auxiliary tasks, which reduce the inference speed. To maintain the efficiency of the sparse framework while fully aggregating contextual information, in this work, we propose SparseDet which designs sparse queries as object proxies. It introduces two key modules, the Local Multi-scale Feature Aggregation (LMFA) module and the Global Feature Aggregation (GFA) module, aiming to fully capture the contextual information, thereby enhancing the ability of the proxies to represent objects. Where LMFA sub-module achieves feature fusion across different scales for sparse key voxels %which does this through via coordinate transformations and using nearest neighbor relationships to capture object-level details and local contextual information, GFA sub-module uses self-attention mechanisms to selectively aggregate the features of the key voxels across the entire scene for capturing scene-level contextual information. Experiments on nuScenes and KITTI demonstrate the effectiveness of our method. Specifically, on nuScene, SparseDet surpasses the previous best sparse detector VoxelNeXt by 2.2\% mAP with 13.5 FPS, and on KITTI, it surpasses VoxelNeXt by 1.12\% $\mathbf{AP_{3D}}$ on hard level tasks with 17.9 FPS.
Abstract:Games are widely used as research environments for multi-agent reinforcement learning (MARL), but they pose three significant challenges: limited customization, high computational demands, and oversimplification. To address these issues, we introduce the first publicly available map editor for the popular mobile game Honor of Kings and design a lightweight environment, Mini Honor of Kings (Mini HoK), for researchers to conduct experiments. Mini HoK is highly efficient, allowing experiments to be run on personal PCs or laptops while still presenting sufficient challenges for existing MARL algorithms. We have tested our environment on common MARL algorithms and demonstrated that these algorithms have yet to find optimal solutions within this environment. This facilitates the dissemination and advancement of MARL methods within the research community. Additionally, we hope that more researchers will leverage the Honor of Kings map editor to develop innovative and scientifically valuable new maps. Our code and user manual are available at: https://github.com/tencent-ailab/mini-hok.
Abstract:We introduce Seed-TTS, a family of large-scale autoregressive text-to-speech (TTS) models capable of generating speech that is virtually indistinguishable from human speech. Seed-TTS serves as a foundation model for speech generation and excels in speech in-context learning, achieving performance in speaker similarity and naturalness that matches ground truth human speech in both objective and subjective evaluations. With fine-tuning, we achieve even higher subjective scores across these metrics. Seed-TTS offers superior controllability over various speech attributes such as emotion and is capable of generating highly expressive and diverse speech for speakers in the wild. Furthermore, we propose a self-distillation method for speech factorization, as well as a reinforcement learning approach to enhance model robustness, speaker similarity, and controllability. We additionally present a non-autoregressive (NAR) variant of the Seed-TTS model, named $\text{Seed-TTS}_\text{DiT}$, which utilizes a fully diffusion-based architecture. Unlike previous NAR-based TTS systems, $\text{Seed-TTS}_\text{DiT}$ does not depend on pre-estimated phoneme durations and performs speech generation through end-to-end processing. We demonstrate that this variant achieves comparable performance to the language model-based variant and showcase its effectiveness in speech editing. We encourage readers to listen to demos at \url{https://bytedancespeech.github.io/seedtts_tech_report}.
Abstract:In the field of 3D object detection tasks, fusing heterogeneous features from LiDAR and camera sensors into a unified Bird's Eye View (BEV) representation is a widely adopted paradigm. However, existing methods are often compromised by imprecise sensor calibration, resulting in feature misalignment in LiDAR-camera BEV fusion. Moreover, such inaccuracies result in errors in depth estimation for the camera branch, ultimately causing misalignment between LiDAR and camera BEV features. In this work, we propose a novel ContrastAlign approach that utilizes contrastive learning to enhance the alignment of heterogeneous modalities, thereby improving the robustness of the fusion process. Specifically, our approach includes the L-Instance module, which directly outputs LiDAR instance features within LiDAR BEV features. Then, we introduce the C-Instance module, which predicts camera instance features through RoI (Region of Interest) pooling on the camera BEV features. We propose the InstanceFusion module, which utilizes contrastive learning to generate similar instance features across heterogeneous modalities. We then use graph matching to calculate the similarity between the neighboring camera instance features and the similarity instance features to complete the alignment of instance features. Our method achieves state-of-the-art performance, with an mAP of 70.3%, surpassing BEVFusion by 1.8% on the nuScenes validation set. Importantly, our method outperforms BEVFusion by 7.3% under conditions with misalignment noise.