arXiv subjects

Yuke Li

Publications and source records attributed to Yuke Li.

At least 73 records · Page 4Linked to original sources

arXiv2024

Split to Merge: Unifying Separated Modalities for Unsupervised Domain Adaptation

Large vision-language models (VLMs) like CLIP have demonstrated good zero-shot learning performance in the unsupervised domain adaptation task. Yet, most transfer approaches for VLMs focus on either the language or visual branches, overlooking the nuanced interplay between both modalities. In this work, we introduce a Unified Modality Separation (UniMoS) framework for unsupervised domain adaptation. Leveraging insights from modality gap studies, we craft a nimble modality separation network that distinctly disentangles CLIP's features into language-associated and vision-associated components. Our proposed Modality-Ensemble Training (MET) method fosters the exchange of modality-agnostic information while maintaining modality-specific nuances. We align features across domains using a modality discriminator. Comprehensive evaluations on three benchmarks reveal our approach sets a new state-of-the-art with minimal computational costs. Code: https://github.com/TL-UESTC/UniMoS

cs.CV↗

arXiv2024

HaltingVT: Adaptive Token Halting Transformer for Efficient Video Recognition

Action recognition in videos poses a challenge due to its high computational cost, especially for Joint Space-Time video transformers (Joint VT). Despite their effectiveness, the excessive number of tokens in such architectures significantly limits their efficiency. In this paper, we propose HaltingVT, an efficient video transformer adaptively removing redundant video patch tokens, which is primarily composed of a Joint VT and a Glimpser module. Specifically, HaltingVT applies data-adaptive token reduction at each layer, resulting in a significant reduction in the overall computational cost. Besides, the Glimpser module quickly removes redundant tokens in shallow transformer layers, which may even be misleading for video recognition tasks based on our observations. To further encourage HaltingVT to focus on the key motion-related information in videos, we design an effective Motion Loss during training. HaltingVT acquires video analysis capabilities and token halting compression strategies simultaneously in a unified training process, without requiring additional training procedures or sub-networks. On the Mini-Kinetics dataset, we achieved 75.0% top-1 ACC with 24.2 GFLOPs, as well as 67.2% top-1 ACC with an extremely low 9.9 GFLOPs. The code is available at https://github.com/dun-research/HaltingVT.

cs.CV↗

arXiv2023

Learning Socio-Temporal Graphs for Multi-Agent Trajectory Prediction

In order to predict a pedestrian's trajectory in a crowd accurately, one has to take into account her/his underlying socio-temporal interactions with other pedestrians consistently. Unlike existing work that represents the relevant information separately, partially, or implicitly, we propose a complete representation for it to be fully and explicitly captured and analyzed. In particular, we introduce a Directed Acyclic Graph-based structure, which we term Socio-Temporal Graph (STG), to explicitly capture pair-wise socio-temporal interactions among a group of people across both space and time. Our model is built on a time-varying generative process, whose latent variables determine the structure of the STGs. We design an attention-based model named STGformer that affords an end-to-end pipeline to learn the structure of the STGs for trajectory prediction. Our solution achieves overall state-of-the-art prediction accuracy in two large-scale benchmark datasets. Our analysis shows that a person's past trajectory is critical for predicting another person's future path. Our model learns this relationship with a strong notion of socio-temporal localities. Statistics show that utilizing this information explicitly for prediction yields a noticeable performance gain with respect to the trajectory-only approaches.

cs.CV↗

arXiv2023

LAE-ST-MoE: Boosted Language-Aware Encoder Using Speech Translation Auxiliary Task for E2E Code-switching ASR

Recently, to mitigate the confusion between different languages in code-switching (CS) automatic speech recognition (ASR), the conditionally factorized models, such as the language-aware encoder (LAE), explicitly disregard the contextual information between different languages. However, this information may be helpful for ASR modeling. To alleviate this issue, we propose the LAE-ST-MoE framework. It incorporates speech translation (ST) tasks into LAE and utilizes ST to learn the contextual information between different languages. It introduces a task-based mixture of expert modules, employing separate feed-forward networks for the ASR and ST tasks. Experimental results on the ASRU 2019 Mandarin-English CS challenge dataset demonstrate that, compared to the LAE-based CTC, the LAE-ST-MoE model achieves a 9.26% mix error reduction on the CS test with the same decoding parameter. Moreover, the well-trained LAE-ST-MoE model can perform ST tasks from CS speech to Mandarin or English text.

cs.SD↗

arXiv2023

Zero-Shot Emotion Transfer For Cross-Lingual Speech Synthesis

Zero-shot emotion transfer in cross-lingual speech synthesis aims to transfer emotion from an arbitrary speech reference in the source language to the synthetic speech in the target language. Building such a system faces challenges of unnatural foreign accents and difficulty in modeling the shared emotional expressions of different languages. Building on the DelightfulTTS neural architecture, this paper addresses these challenges by introducing specifically-designed modules to model the language-specific prosody features and language-shared emotional expressions separately. Specifically, the language-specific speech prosody is learned by a non-autoregressive predictive coding (NPC) module to improve the naturalness of the synthetic cross-lingual speech. The shared emotional expression between different languages is extracted from a pre-trained self-supervised model HuBERT with strong generalization capabilities. We further use hierarchical emotion modeling to capture more comprehensive emotions across different languages. Experimental results demonstrate the proposed framework's effectiveness in synthesizing bi-lingual emotional speech for the monolingual target speaker without emotional training data.

cs.SD↗

arXiv2023

HPL-ViT: A Unified Perception Framework for Heterogeneous Parallel LiDARs in V2V

To develop the next generation of intelligent LiDARs, we propose a novel framework of parallel LiDARs and construct a hardware prototype in our experimental platform, DAWN (Digital Artificial World for Natural). It emphasizes the tight integration of physical and digital space in LiDAR systems, with networking being one of its supported core features. In the context of autonomous driving, V2V (Vehicle-to-Vehicle) technology enables efficient information sharing between different agents which significantly promotes the development of LiDAR networks. However, current research operates under an ideal situation where all vehicles are equipped with identical LiDAR, ignoring the diversity of LiDAR categories and operating frequencies. In this paper, we first utilize OpenCDA and RLS (Realistic LiDAR Simulation) to construct a novel heterogeneous LiDAR dataset named OPV2V-HPL. Additionally, we present HPL-ViT, a pioneering architecture designed for robust feature fusion in heterogeneous and dynamic scenarios. It uses a graph-attention Transformer to extract domain-specific features for each agent, coupled with a cross-attention mechanism for the final fusion. Extensive experiments on OPV2V-HPL demonstrate that HPL-ViT achieves SOTA (state-of-the-art) performance in all settings and exhibits outstanding generalization capabilities.

cs.CV↗

arXiv2023

Part-Aware Transformer for Generalizable Person Re-identification

Domain generalization person re-identification (DG-ReID) aims to train a model on source domains and generalize well on unseen domains. Vision Transformer usually yields better generalization ability than common CNN networks under distribution shifts. However, Transformer-based ReID models inevitably over-fit to domain-specific biases due to the supervised learning strategy on the source domain. We observe that while the global images of different IDs should have different features, their similar local parts (e.g., black backpack) are not bounded by this constraint. Motivated by this, we propose a pure Transformer model (termed Part-aware Transformer) for DG-ReID by designing a proxy task, named Cross-ID Similarity Learning (CSL), to mine local visual information shared by different IDs. This proxy task allows the model to learn generic features because it only cares about the visual similarity of the parts regardless of the ID labels, thus alleviating the side effect of domain-specific biases. Based on the local similarity obtained in CSL, a Part-guided Self-Distillation (PSD) is proposed to further improve the generalization of global features. Our method achieves state-of-the-art performance under most DG ReID settings. Under the Market$\to$Duke setting, our method exceeds state-of-the-art by 10.9% and 12.8% in Rank1 and mAP, respectively. The code is available at https://github.com/liyuke65535/Part-Aware-Transformer.

cs.CV↗

arXiv2023

Differentiable Resolution Compression and Alignment for Efficient Video Classification and Retrieval

Optimizing video inference efficiency has become increasingly important with the growing demand for video analysis in various fields. Some existing methods achieve high efficiency by explicit discard of spatial or temporal information, which poses challenges in fast-changing and fine-grained scenarios. To address these issues, we propose an efficient video representation network with Differentiable Resolution Compression and Alignment mechanism, which compresses non-essential information in the early stage of the network to reduce computational costs while maintaining consistent temporal correlations. Specifically, we leverage a Differentiable Context-aware Compression Module to encode the saliency and non-saliency frame features, refining and updating the features into a high-low resolution video sequence. To process the new sequence, we introduce a new Resolution-Align Transformer Layer to capture global temporal correlations among frame features with different resolutions, while reducing spatial computation costs quadratically by utilizing fewer spatial tokens in low-resolution non-saliency frames. The entire network can be end-to-end optimized via the integration of the differentiable compression module. Experimental results show that our method achieves the best trade-off between efficiency and performance on near-duplicate video retrieval and competitive results on dynamic video classification compared to state-of-the-art methods. Code:https://github.com/dun-research/DRCA

cs.CV↗

arXiv2023

Language-Routing Mixture of Experts for Multilingual and Code-Switching Speech Recognition

Multilingual speech recognition for both monolingual and code-switching speech is a challenging task. Recently, based on the Mixture of Experts (MoE), many works have made good progress in multilingual and code-switching ASR, but present huge computational complexity with the increase of supported languages. In this work, we propose a computation-efficient network named Language-Routing Mixture of Experts (LR-MoE) for multilingual and code-switching ASR. LR-MoE extracts language-specific representations through the Mixture of Language Experts (MLE), which is guided to learn by a frame-wise language routing mechanism. The weight-shared frame-level language identification (LID) network is jointly trained as the shared pre-router of each MoE layer. Experiments show that the proposed method significantly improves multilingual and code-switching speech recognition performances over baseline with comparable computational efficiency.

cs.SD↗

arXiv2023

Enhancing the Unified Streaming and Non-streaming Model with Contrastive Learning

The unified streaming and non-streaming speech recognition model has achieved great success due to its comprehensive capabilities. In this paper, we propose to improve the accuracy of the unified model by bridging the inherent representation gap between the streaming and non-streaming modes with a contrastive objective. Specifically, the top-layer hidden representation at the same frame of the streaming and non-streaming modes are regarded as a positive pair, encouraging the representation of the streaming mode close to its non-streaming counterpart. The multiple negative samples are randomly selected from the rest frames of the same sample under the non-streaming mode. Experimental results demonstrate that the proposed method achieves consistent improvements toward the unified model in both streaming and non-streaming modes. Our method achieves CER of 4.66% in the streaming mode and CER of 4.31% in the non-streaming mode, which sets a new state-of-the-art on the AISHELL-1 benchmark.

cs.CL↗

arXiv2023

Improving CTC-based ASR Models with Gated Interlayer Collaboration

The CTC-based automatic speech recognition (ASR) models without the external language model usually lack the capacity to model conditional dependencies and textual interactions. In this paper, we present a Gated Interlayer Collaboration (GIC) mechanism to improve the performance of CTC-based models, which introduces textual information into the model and thus relaxes the conditional independence assumption of CTC-based models. Specifically, we consider the weighted sum of token embeddings as the textual representation for each position, where the position-specific weights are the softmax probability distribution constructed via inter-layer auxiliary CTC losses. The textual representations are then fused with acoustic features by developing a gate unit. Experiments on AISHELL-1, TEDLIUM2, and AIDATATANG corpora show that the proposed method outperforms several strong baselines.

cs.CL↗

arXiv2023

Pressure tuning domain-wall chirality in noncentrosymmetric magnetic Weyl semimetal CeAlGe

Topological magnetic Weyl semimetals have been proposed to host controllable chiral domain walls which bear a great prospect in device applications. To exploit them in applications, it is important to have a proper way to tune and manipulate these domain walls. One possible means is through magnetoelastic coupling. The involvement of rare earth in the lately proposed $R$Al$X$ ($R$=rare earth, $X$=Si and Ge) family magnetic Weyl semimetals may provide such a platform. Here we present transport and thermodynamic properties of CeAlGe under hydrostatic pressure. We find that pressure enhances the antiferromagnetic exchange in CeAlGe but essentially retains its magnetic structure. Large topological Hall effect with pronounced loop shape is observed within the magnetically ordered state, and it splits into two regions under pressure. Such an unusual electromagnetic response is inferred to be a consequence of chiral magnetic domain walls. The unprecedented concomitance of its evolution under pressure and the reentrance of antiferromagnetic order strongly suggest the capability of switching on/off this electromagnetic response in noncentrosymmetric magnetic Weyl semimetals via magnetoelastic coupling.

cond-mat.str-el↗

arXiv2022

Acoustic Pornography Recognition Using Convolutional Neural Networks and Bag of Refinements

A large number of pornographic audios publicly available on the Internet seriously threaten the mental and physical health of children, but these audios are rarely detected and filtered. In this paper, we firstly propose a convolutional neural networks (CNN) based model for acoustic pornography recognition. Then, we research a collection of refinements and verify their effectiveness through ablation studies. Finally, we stack all refinements together to verify whether they can further improve the accuracy of the model. Experimental results on our newly-collected large dataset consisting of 224127 pornographic audios and 274206 normal samples demonstrate the effectiveness of our proposed model and these refinements. Specifically, the proposed model achieves an accuracy of 92.46% and the accuracy is further improved to 97.19% when all refinements are combined.

cs.SD↗

arXiv2022

3D-CSL: self-supervised 3D context similarity learning for Near-Duplicate Video Retrieval

In this paper, we introduce 3D-CSL, a compact pipeline for Near-Duplicate Video Retrieval (NDVR), and explore a novel self-supervised learning strategy for video similarity learning. Most previous methods only extract video spatial features from frames separately and then design kinds of complex mechanisms to learn the temporal correlations among frame features. However, parts of spatiotemporal dependencies have already been lost. To address this, our 3D-CSL extracts global spatiotemporal dependencies in videos end-to-end with a 3D transformer and find a good balance between efficiency and effectiveness by matching on clip-level. Furthermore, we propose a two-stage self-supervised similarity learning strategy to optimize the entire network. Firstly, we propose PredMAE to pretrain the 3D transformer with video prediction task; Secondly, ShotMix, a novel video-specific augmentation, and FCS loss, a novel triplet loss, are proposed further promote the similarity learning results. The experiments on FIVR-200K and CC_WEB_VIDEO demonstrate the superiority and reliability of our method, which achieves the state-of-the-art performance on clip-level NDVR.

cs.CV↗

arXiv2022

Multi-Level Modeling Units for End-to-End Mandarin Speech Recognition

The choice of modeling units is crucial for automatic speech recognition (ASR) tasks. In mandarin scenarios, the Chinese characters represent meaning but are not directly related to the pronunciation. Thus only considering the writing of Chinese characters as modeling units is insufficient to capture speech features. In this paper, we present a novel method involves with multi-level modeling units, which integrates multi-level information for mandarin speech recognition. Specifically, the encoder block considers syllables as modeling units and the decoder block deals with character-level modeling units. To facilitate the incremental conversion from syllable features to character features, we design an auxiliary task that applies cross-entropy (CE) loss to intermediate decoder layers. During inference, the input feature sequences are converted into syllable sequences by the encoder block and then converted into Chinese characters by the decoder block. Experiments on the widely used AISHELL-1 corpus demonstrate that our method achieves promising results with CER of 4.1%/4.6% and 4.6%/5.2%, using the Conformer and the Transformer backbones respectively.

cs.CL↗

arXiv2022

Anomalous ferromagnetic behavior in the orthorhombic Li$_3$Co$_2$SbO$_6$

Monoclinic Li$_3$Co$_2$SbO$_6$ has been proposed as a Kitaev spin liquid candidate and investigated intensively, whereas the properties of its polymorph, the orthorhombic phase, is less known. Here we report the magnetic properties of the orthorhombic Li$_3$Co$_2$SbO$_6$ as revealed by dc and ac magnetic susceptibility, muon spin relaxation ($μ$SR) and neutron diffraction measurements. Successive magnetic transitions at (115, 89 and 71) K were observed in the low field dc susceptibility measurements. The transitions below $T_N$ (= 115 K), are suppressed in higher applied fields. However, zero field, ac susceptibility measurements reveals distinct frequency independent transitions at about (114, 107, 97, 79 and 71) K. A long range magnetic ordered state was confirmed by specific heat, $μ$SR and neutron diffraction measurements, all indicating a single transition at about 115 K. The discrepancy between different measurements is attributed to possible stacking faults and/or local disorders of the ferromagnetic zig-zag chains, resulting in ferromagnetic boundaries within the overall antiferromagnetic matrix.

cond-mat.str-el↗

arXiv2022

Sign-reversed anomalous Nernst effect in the ferromagnetic Weyl-semimetal Fe$_{3-x}$GeTe$_2$: the role of Fe vacancies

Anomalous Nernst effect, as a thermal partner of anomalous Hall effect, is particularly sensitive to the Berry curvature anomaly near the Fermi level, and has been used to probe the topological nature of quantum materials. In this work, we report the observation of both effects in the ferromagnetic Weyl-semimetal Fe$_{3-x}$GeTe$_2$ with tunable Fe vacancies. With decreasing Fe vacancies, the anomalous Hall conductivity evolves as a function of the longitudinal conductivity from the hopping region to the region where the intrinsic Berry curvature contribution dominates. Concomitant evolutions in the anomalous Nernst signal and the anomalous off-diagonal thermoelectric coefficient are observed below the Curie temperature, displaying a unique sign change caused by the Fe vacancies. Combining these results with first-principles calculations, we argue that the Fe-vacancy concentration plays a unique role in simultaneously tuning the chemical potential and ferromagnetism, which in turn controls the Berry curvature contribution in this family of ferromagnetic topological semimetals.

cond-mat.mtrl-sci↗

arXiv2022

Elastic instability in a straight channel of viscoelastic flow without prearranged perturbations

We report experimental results on elastic instability in a viscoelastic channel shear flow due to only a natural non-smoothed inlet and small holes along the channel for pressure measurements. We show that non-normal mode instability results in elastic waves and chaotic flow self-organized into periodically cycled stream-wise streaks synchronized by elastic wave frequency. The chaotic flow persists above the transition with increasing $Wi$ further into elastic turbulence and drag reduction regimes. Thus, we resolve the recent puzzle whether strong prearranged perturbations are necessary to get an elastic instability in parallel shear viscoelastic flow. Moreover, flow resistance, velocity spectra decay, and elastic wave speed reveal the same scaling with Wi as obtained in the case of strong disturbances. The remarkable result is that all scaling behavior and streaks are found in the entire channel with small attenuation, in sharp contrast to the flow with strong prearranged perturbations.

physics.flu-dyn↗