Wei Zou

Hu Su, Yonghao He, Rui Jiang et al.

Anchor-free detectors basically formulate object detection as dense classification and regression. For popular anchor-free detectors, it is common to introduce an individual prediction branch to estimate the quality of localization. The following inconsistencies are observed when we delve into the practices of classification and quality estimation. Firstly, for some adjacent samples which are assigned completely different labels, the trained model would produce similar classification scores. This violates the training objective and leads to performance degradation. Secondly, it is found that detected bounding boxes with higher confidences contrarily have smaller overlaps with the corresponding ground-truth. Accurately localized bounding boxes would be suppressed by less accurate ones in the Non-Maximum Suppression (NMS) procedure. To address the inconsistency problems, the Dynamic Smooth Label Assignment (DSLA) method is proposed. Based on the concept of centerness originally developed in FCOS, a smooth assignment strategy is proposed. The label is smoothed to a continuous value in [0, 1] to make a steady transition between positive and negative samples. Intersection-of-Union (IoU) is predicted dynamically during training and is coupled with the smoothed label. The dynamic smooth label is assigned to supervise the classification branch. Under such supervision, quality estimation branch is naturally merged into the classification branch, which simplifies the architecture of anchor-free detector. Comprehensive experiments are conducted on the MS COCO benchmark. It is demonstrated that, DSLA can significantly boost the detection accuracy by alleviating the above inconsistencies for anchor-free detectors. Our codes are released at https://github.com/YonghaoHe/DSLA.

Cheng Wen, Xianghui Sun, Shuaijiang Zhao et al.

This paper presents the development and evaluation of ChatHome, a domain-specific language model (DSLM) designed for the intricate field of home renovation. Considering the proven competencies of large language models (LLMs) like GPT-4 and the escalating fascination with home renovation, this study endeavors to reconcile these aspects by generating a dedicated model that can yield high-fidelity, precise outputs relevant to the home renovation arena. ChatHome's novelty rests on its methodology, fusing domain-adaptive pretraining and instruction-tuning over an extensive dataset. This dataset includes professional articles, standard documents, and web content pertinent to home renovation. This dual-pronged strategy is designed to ensure that our model can assimilate comprehensive domain knowledge and effectively address user inquiries. Via thorough experimentation on diverse datasets, both universal and domain-specific, including the freshly introduced "EvalHome" domain dataset, we substantiate that ChatHome not only amplifies domain-specific functionalities but also preserves its versatility.

Huan Shen, Yingao Wang, Shangkun Huang et al.

Turn-taking modeling is fundamental to spoken dialogue systems, yet its evaluation remains fragmented and often limited to binary boundary detection under narrow interaction settings. Such protocols hinder systematic comparison and obscure model weaknesses across conversational conditions. We present CoDeTT, a context-aware decision benchmark for turn-taking evaluation. CoDeTT formulates turn-taking as a structured decision problem and constructs a multi-scenario dataset with fine-grained decision categories and controlled context variations. Under a unified evaluation protocol, we assess representative existing models and observe substantial performance disparities across decision types and interaction scenarios. CoDeTT provides a standardized benchmark for systematic and context-aware evaluation of turn-taking systems. The benchmark dataset and evaluation toolkit are available at https://github.com/YingaoWang-casia/CoDeTT.github.io.

Xiaoyu Tian, Liangyu Chen, Na Liu et al.

Inspired by the dual-process theory of human cognition, we introduce DUMA, a novel conversational agent framework that embodies a dual-mind mechanism through the utilization of two generative Large Language Models (LLMs) dedicated to fast and slow thinking respectively. The fast thinking model serves as the primary interface for external interactions and initial response generation, evaluating the necessity for engaging the slow thinking model based on the complexity of the complete response. When invoked, the slow thinking model takes over the conversation, engaging in meticulous planning, reasoning, and tool utilization to provide a well-analyzed response. This dual-mind configuration allows for a seamless transition between intuitive responses and deliberate problem-solving processes based on the situation. We have constructed a conversational agent to handle online inquiries in the real estate industry. The experiment proves that our method balances effectiveness and efficiency, and has a significant improvement compared to the baseline.

Goutham Rajendran, Wei Zou

We investigate robustness properties of pre-trained neural models for automatic speech recognition. Real life data in machine learning is usually very noisy and almost never clean, which can be attributed to various factors depending on the domain, e.g. outliers, random noise and adversarial noise. Therefore, the models we develop for various tasks should be robust to such kinds of noisy data, which led to the thriving field of robust machine learning. We consider this important issue in the setting of automatic speech recognition. With the increasing popularity of pre-trained models, it's an important question to analyze and understand the robustness of such models to noise. In this work, we perform a robustness analysis of the pre-trained neural models wav2vec2, HuBERT and DistilHuBERT on the LibriSpeech and TIMIT datasets. We use different kinds of noising mechanisms and measure the model performances as quantified by the inference time and the standard Word Error Rate metric. We also do an in-depth layer-wise analysis of the wav2vec2 model when injecting noise in between layers, enabling us to predict at a high level what each layer learns. Finally for this model, we visualize the propagation of errors across the layers and compare how it behaves on clean versus noisy data. Our experiments conform the predictions of Pasad et al. [2021] and also raise interesting directions for future work.

Weifei Jin, Xilong Wang, Wei Zou et al.

Retrieval-augmented generation (RAG) is vulnerable to prompt injection attacks, in which an adversary inserts malicious documents containing carefully crafted injected prompts into the knowledge database. When a user issues a question targeted by the attack, the RAG system may retrieve these malicious documents, whose injected prompts mislead it into generating attacker-specified answers, thereby compromising the integrity of the RAG system. In this work, we propose CleanBase, a method to detect malicious documents within a knowledge database. Our key insight is that malicious documents crafted for the same attack-targeted questions often exhibit high semantic similarity, as attackers deliberately make them consistent to improve attack success rates. Accordingly, CleanBase constructs a similarity graph over the knowledge database, where each node represents a document and an edge connects two nodes if their semantic similarity--computed using an embedding model--exceeds a statistically determined threshold. Due to their inherent similarity, malicious documents tend to form cliques within this graph. CleanBase detects such cliques and flags the corresponding documents as malicious. We theoretically derive upper bounds on CleanBase's false positive and false negative rates and empirically validate its effectiveness. Experimental results across multiple datasets and prompt injection attacks demonstrate that CleanBase accurately detects malicious documents and effectively safeguards RAG systems. Our source code is available at https://github.com/WeifeiJin/CleanBase.

Jiaojiao Guan, Yongxin Ji, Cheng Peng et al.

Bacteriophages are viruses that target bacteria, playing a crucial role in microbial ecology. Phage proteins are important in understanding phage biology, such as virus infection, replication, and evolution. Although a large number of new phages have been identified via metagenomic sequencing, many of them have limited protein function annotation. Accurate function annotation of phage proteins presents several challenges, including their inherent diversity and the scarcity of annotated ones. Existing tools have yet to fully leverage the unique properties of phages in annotating protein functions. In this work, we propose a new protein function annotation tool for phages by leveraging the modular genomic structure of phage genomes. By employing embeddings from the latest protein foundation models and Transformer to capture contextual information between proteins in phage genomes, PhaGO surpasses state-of-the-art methods in annotating diverged proteins and proteins with uncommon functions by 6.78% and 13.05% improvement, respectively. PhaGO can annotate proteins lacking homology search results, which is critical for characterizing the rapidly accumulating phage genomes. We demonstrate the utility of PhaGO by identifying 688 potential holins in phages, which exhibit high structural conservation with known holins. The results show the potential of PhaGO to extend our understanding of newly discovered phages.

Xiaoyu Fan, Huizhi Xie, Wei Zou et al.

Large language model (LLM)-based text-to-speech (TTS) systems achieve remarkable naturalness via autoregressive (AR) decoding, but require N sequential steps to generate N speech tokens. We present LLaDA-TTS, which replaces the AR LLM with a masked diffusion model that completes generation in a fixed number of parallel steps, decoupling inference latency from sequence length. Remarkably, using only 50 hours of fine-tuning data, we successfully transfer a pretrained AR checkpoint to the masked diffusion paradigm via bidirectional attention. At 64 steps, LLaDA-TTS achieves 0.98% CER (zh) and 1.96% WER (en) on Seed-TTS-Eval, matching the original CosyVoice 3 baseline performance while delivering a 2x LLM-stage speedup--a notable acceleration achieved despite the absence of KV cache, an optimization the AR baseline heavily relies on. Beyond acceleration, the bidirectional architecture naturally enables zero-shot speech editing--including word-level insertion, deletion, and substitution--without any additional training. Theoretically, we prove that AR-pretrained weights are near-optimal for bidirectional masked prediction under the locality property of acoustic tokens, explaining this rapid convergence. This general method modifies only the attention mask and objective, applying seamlessly to any LLM-based AR TTS system. Code and audio samples will be available at https://deft-piroshki-b652b5.netlify.app/.

Xiang Geng, Ming Zhu, Jiahuan Li et al.

The scarcity of non-English data limits the development of non-English large language models (LLMs). Transforming English-centric LLMs to non-English has been identified as an effective and resource-efficient method. Previous works start from base LLMs and perform knowledge distillation (KD) with data generated by stronger LLMs, e.g. GPT-4. Compared to base LLMs, chat LLMs are further optimized for advanced abilities, e.g. multi-turn conversation and human preference alignment, and thus more powerful in both helpfulness and safety. However, transforming a chat LLM involves two critical issues: (1) How can we effectively transfer advanced abilities without their supervised data? (2) How can we prevent the original knowledge from catastrophic forgetting during transformation? We target these issues by introducing a simple framework called TransLLM. For the first issue, TransLLM divides the transfer problem into some common sub-tasks with the translation chain-of-thought, which uses the translation as the bridge between English and non-English step-by-step. We further enhance the performance of sub-tasks with publicly available data. For the second issue, we propose a method comprising two synergistic components: low-rank adaptation for training to maintain the original LLM parameters, and recovery KD, which utilizes data generated by the chat LLM itself to recover the original knowledge from the frozen parameters. In the experiments, we transform the LLaMA-2-chat-7B to the Thai language. Our method, using only single-turn data, outperforms strong baselines and ChatGPT on multi-turn benchmark MT-bench. Furthermore, our method, without safety data, rejects more harmful queries of safety benchmark AdvBench than both ChatGPT and GPT-4. Code is available at https://github.com/hy5468/TransLLM.

Shuaijiang Zhao, Tingwei Guo, Bajian Xiang et al.

The GPT-4o represents a significant milestone in enabling real-time interaction with large language models (LLMs) through speech, its remarkable low latency and high fluency not only capture attention but also stimulate research interest in the field. This real-time speech interaction is particularly valuable in scenarios requiring rapid feedback and immediate responses, dramatically enhancing user experience. However, there is a notable lack of research focused on real-time large speech language models, particularly for Chinese. In this work, we present KE-Omni, a seamless large speech language model built upon Ke-SpeechChat, a large-scale high-quality synthetic speech interaction dataset consisting of 7 million Chinese and English conversations, featuring 42,002 speakers, and totaling over 60,000 hours, This contributes significantly to the advancement of research and development in this field. The demos can be accessed at \url{https://huggingface.co/spaces/KE-Team/KE-Omni}.

Wei Zou, Mingwen Dong, Miguel Romero Calvo et al.

Memory makes LLM-based web agents personalized, powerful, yet exploitable. By storing past interactions to personalize future tasks, agents inadvertently create a persistent attack surface that spans websites and sessions. While existing security research on memory assumes attackers can directly inject into memory storage or exploit shared memory across users, we present a more realistic threat model: contamination through environmental observation alone. We introduce Environment-injected Trajectory-based Agent Memory Poisoning (eTAMP), the first attack to achieve cross-session, cross-site compromise without requiring direct memory access. A single contaminated observation (e.g., viewing a manipulated product page) silently poisons an agent's memory and activates during future tasks on different websites, bypassing permission-based defenses. Our experiments on (Visual)WebArena reveal two key findings. First, eTAMP achieves substantial attack success rates: up to 32.5% on GPT-5-mini, 23.4% on GPT-5.2, and 19.5% on GPT-OSS-120B. Second, we discover Frustration Exploitation: agents under environmental stress become dramatically more susceptible, with ASR increasing up to 8 times when agents struggle with dropped clicks or garbled text. Notably, more capable models are not more secure. GPT-5.2 shows substantial vulnerability despite superior task performance. With the rise of AI browsers like OpenClaw, ChatGPT Atlas, and Perplexity Comet, our findings underscore the urgent need for defenses against environment-injected memory poisoning.

Yanting Wang, Wei Zou, Runpeng Geng et al.

Large language models (LLMs) and their applications, such as agents, are highly vulnerable to prompt injection attacks. State-of-the-art prompt injection detection methods have the following limitations: (1) their effectiveness degrades significantly as context length increases, and (2) they lack explicit rules that define what constitutes prompt injection, causing detection decisions to be implicit, opaque, and difficult to reason about. In this work, we propose AgentWatcher to address the above two limitations. To address the first limitation, AgentWatcher attributes the LLM's output (e.g., the action of an agent) to a small set of causally influential context segments. By focusing detection on a relatively short text, AgentWatcher can be scalable to long contexts. To address the second limitation, we define a set of rules specifying what does and does not constitute a prompt injection, and use a monitor LLM to reason over these rules based on the attributed text, making the detection decisions more explainable. We conduct a comprehensive evaluation on tool-use agent benchmarks and long-context understanding datasets. The experimental results demonstrate that AgentWatcher can effectively detect prompt injection and maintain utility without attacks. The code is available at https://github.com/wang-yanting/AgentWatcher.

Yuxin Guo, Teng Wang, Yuying Ge et al.

Recent advances in text-to-audio (TTA) generation excel at synthesizing short audio clips but struggle with long-form narrative audio, which requires temporal coherence and compositional reasoning. To address this gap, we propose AudioStory, a unified framework that integrates large language models (LLMs) with TTA systems to generate structured, long-form audio narratives. AudioStory possesses strong instruction-following reasoning generation capabilities. It employs LLMs to decompose complex narrative queries into temporally ordered sub-tasks with contextual cues, enabling coherent scene transitions and emotional tone consistency. AudioStory has two appealing features: (1) Decoupled bridging mechanism: AudioStory disentangles LLM-diffuser collaboration into two specialized components, i.e., a bridging query for intra-event semantic alignment and a residual query for cross-event coherence preservation. (2) End-to-end training: By unifying instruction comprehension and audio generation within a single end-to-end framework, AudioStory eliminates the need for modular training pipelines while enhancing synergy between components. Furthermore, we establish a benchmark AudioStory-10K, encompassing diverse domains such as animated soundscapes and natural sound narratives. Extensive experiments show the superiority of AudioStory on both single-audio generation and narrative audio generation, surpassing prior TTA baselines in both instruction-following ability and audio fidelity. Our code is available at https://github.com/TencentARC/AudioStory

Yanting Wang, Wei Zou, Runpeng Geng et al.

Long context large language models (LLMs) are deployed in many real-world applications such as RAG, agent, and broad LLM-integrated applications. Given an instruction and a long context (e.g., documents, PDF files, webpages), a long context LLM can generate an output grounded in the provided context, aiming to provide more accurate, up-to-date, and verifiable outputs while reducing hallucinations and unsupported claims. This raises a research question: how to pinpoint the texts (e.g., sentences, passages, or paragraphs) in the context that contribute most to or are responsible for the generated output by an LLM? This process, which we call context traceback, has various real-world applications, such as 1) debugging LLM-based systems, 2) conducting post-attack forensic analysis for attacks (e.g., prompt injection attack, knowledge corruption attacks) to an LLM, and 3) highlighting knowledge sources to enhance the trust of users towards outputs generated by LLMs. When applied to context traceback for long context LLMs, existing feature attribution methods such as Shapley have sub-optimal performance and/or incur a large computational cost. In this work, we develop TracLLM, the first generic context traceback framework tailored to long context LLMs. Our framework can improve the effectiveness and efficiency of existing feature attribution methods. To improve the efficiency, we develop an informed search based algorithm in TracLLM. We also develop contribution score ensemble/denoising techniques to improve the accuracy of TracLLM. Our evaluation results show TracLLM can effectively identify texts in a long context that lead to the output of an LLM. Our code and data are at: https://github.com/Wang-Yanting/TracLLM.

Jiangyuan Liu, Hongxuan Ma, Yuxin Guo et al.

Transparent object perception is indispensable for numerous robotic tasks. However, accurately segmenting and estimating the depth of transparent objects remain challenging due to complex optical properties. Existing methods primarily delve into only one task using extra inputs or specialized sensors, neglecting the valuable interactions among tasks and the subsequent refinement process, leading to suboptimal and blurry predictions. To address these issues, we propose a monocular framework, which is the first to excel in both segmentation and depth estimation of transparent objects, with only a single-image input. Specifically, we devise a novel semantic and geometric fusion module, effectively integrating the multi-scale information between tasks. In addition, drawing inspiration from human perception of objects, we further incorporate an iterative strategy, which progressively refines initial features for clearer results. Experiments on two challenging synthetic and real-world datasets demonstrate that our model surpasses state-of-the-art monocular, stereo, and multi-view methods by a large margin of about 38.8%-46.2% with only a single RGB input. Codes and models are publicly available at https://github.com/L-J-Yuan/MODEST.

Jiangyuan Liu, Hongxuan Ma, Yuhao Zhao et al.

Point cloud completion aims to recover complete 3D geometry from partial observations caused by limited viewpoints and occlusions. Existing learning-based works, including 3D Convolutional Neural Network (CNN)-based, point-based, and Transformer-based methods, have achieved strong performance on synthetic benchmarks. However, due to the limitations of modality, scalability, and generative capacity, their generalization to novel objects and real-world scenarios remains challenging. In this paper, we propose MGPC, a generalizable multimodal point cloud completion framework that integrates point clouds, RGB images, and text within a unified architecture. MGPC introduces an innovative modality dropout strategy, a Transformer-based fusion module, and a novel progressive generator to improve robustness, scalability, and geometric modeling capability. We further develop an automatic data generation pipeline and construct MGPC-1M, a large-scale benchmark with over 1,000 categories and one million training pairs. Extensive experiments on MGPC-1M and in-the-wild data demonstrate that the proposed method consistently outperforms prior baselines and exhibits strong generalization under real-world conditions.

Ziqian Wang, Yonghao He, Licheng Yang et al.

Generating high-fidelity, physically interactive 3D simulated tabletop scenes is essential for embodied AI--especially for robotic manipulation policy learning and data synthesis. However, current text- or image-driven 3D scene generation methods mainly focus on large-scale scenes, struggling to capture the high-density layouts and complex spatial relations that characterize tabletop scenes. To address these challenges, we propose TabletopGen, a training-free, fully automatic framework that generates diverse, instance-level interactive 3D tabletop scenes. TabletopGen accepts a reference image as input, which can be synthesized by a text-to-image model to enhance scene diversity. We then perform instance segmentation and completion on the reference to obtain per-instance images. Each instance is reconstructed into a 3D model followed by canonical coordinate alignment. The aligned 3D models then undergo pose and scale estimation before being assembled into a collision-free, simulation-ready tabletop scene. A key component of our framework is a novel pose and scale alignment approach that decouples the complex spatial reasoning into two stages: a Differentiable Rotation Optimizer for precise rotation recovery and a Top-view Spatial Alignment mechanism for robust translation and scale estimation, enabling accurate 3D reconstruction from 2D reference. Extensive experiments and user studies show that TabletopGen achieves state-of-the-art performance, markedly surpassing existing methods in visual fidelity, layout accuracy, and physical plausibility, capable of generating realistic tabletop scenes with rich stylistic and spatial diversity. Our code will be publicly available.

Yu Kang, Xianghui Sun, Liangyu Chen et al.

Generating Chain-of-Thought (CoT) before deriving the answer can effectively improve the reasoning capabilities of large language models (LLMs) and significantly improve the accuracy of the generated answer. However, in most cases, the length of the generated CoT is much longer than the desired final answer, which results in additional decoding costs. Furthermore, existing research has discovered that shortening the reasoning steps in CoT, even while preserving the key information, diminishes LLMs' abilities. These phenomena make it difficult to use LLMs and CoT in many real-world applications that only require the final answer and are sensitive to latency, such as search and recommendation. To reduce the costs of model decoding and shorten the length of the generated CoT, this paper presents $\textbf{C}$onditioned $\textbf{C}$ompressed $\textbf{C}$hain-of-$\textbf{T}$hought (C3oT), a CoT compression framework that involves a compressor to compress an original longer CoT into a shorter CoT while maintaining key information and interpretability, a conditioned training method to train LLMs with both longer CoT and shorter CoT simultaneously to learn the corresponding relationships between them, and a conditioned inference method to gain the reasoning ability learned from longer CoT by generating shorter CoT. We conduct experiments over four datasets from arithmetic and commonsense scenarios, showing that the proposed method is capable of compressing the length of generated CoT by up to more than 50% without compromising its effectiveness.

Peng Ding, Wen Sun, Dailin Li et al.

Large Language Models (LLMs) excel at various natural language processing tasks but remain vulnerable to jailbreaking attacks that induce harmful content generation. In this paper, we reveal a critical safety inconsistency: LLMs can more effectively identify harmful requests as discriminators than defend against them as generators. This insight inspires us to explore aligning the model's inherent discrimination and generation capabilities. To this end, we propose SDGO (Self-Discrimination-Guided Optimization), a reinforcement learning framework that leverages the model's own discrimination capabilities as a reward signal to enhance generation safety through iterative self-improvement. Our method does not require any additional annotated data or external models during the training phase. Extensive experiments demonstrate that SDGO significantly improves model safety compared to both prompt-based and training-based baselines while maintaining helpfulness on general benchmarks. By aligning LLMs' discrimination and generation capabilities, SDGO brings robust performance against out-of-distribution (OOD) jailbreaking attacks. This alignment achieves tighter coupling between these two capabilities, enabling the model's generation capability to be further enhanced with only a small amount of discriminative samples. Our code and datasets are available at https://github.com/NJUNLP/SDGO.

Zhaobing Kang, Wei Zou, Zheng Zhu et al.

This paper presents a generic 6DOF camera pose estimation method, which can be used for both the pinhole camera and the fish-eye camera. Different from existing methods, relative positions of 3D points rather than absolute coordinates in the world coordinate system are employed in our method, and it has a unique solution. The application scope of POSIT (Pose from Orthography and Scaling with Iteration) algorithm is generalized to fish-eye cameras by combining with the radially symmetric projection model. The image point relationship between the pinhole camera and the fish-eye camera is derived based on their projection model. The general pose expression which fits for different cameras can be acquired by four noncoplanar object points and their corresponding image points. Accurate estimation results are calculated iteratively. Experimental results on synthetic and real data show that the pose estimation results of our method are more stable and accurate than state-of-the-art methods. The source code is available at https://github.com/k032131/EPOSIT.

Shangkun Huang, Huan Shen, Wei Zou et al.

Speech LLM-based ASR often struggles with named entities and long-tail words due to strong internal language-model priors. Retrieval-augmented biasing can help, but its effectiveness depends on accurate hotword localization in full-utterance speech under weak supervision. We propose CLAR, a dual-encoder speech-text retriever that uses Continuous Integrate-and-Fire (CIF) to learn monotonic token-level alignments without timestamps. With length-aware localized matching, CLAR anchors short-entity acoustic cues and reduces representation dilution and attention drift. The retriever is trained with a multi-granularity objective combining global and local segment-level contrastive losses and a CIF quantity constraint. At inference, top-ranked hotwords are injected as contextual prompts for the Speech LLM, improving recognition without shallow fusion. Experiments show that CLAR significantly improves hotword retrieval and reduces both CER and B-WER against strong contextual ASR baselines.

Wei Zou, Runpeng Geng, Binghui Wang et al.

Large language models (LLMs) have achieved remarkable success due to their exceptional generative capabilities. Despite their success, they also have inherent limitations such as a lack of up-to-date knowledge and hallucination. Retrieval-Augmented Generation (RAG) is a state-of-the-art technique to mitigate these limitations. The key idea of RAG is to ground the answer generation of an LLM on external knowledge retrieved from a knowledge database. Existing studies mainly focus on improving the accuracy or efficiency of RAG, leaving its security largely unexplored. We aim to bridge the gap in this work. We find that the knowledge database in a RAG system introduces a new and practical attack surface. Based on this attack surface, we propose PoisonedRAG, the first knowledge corruption attack to RAG, where an attacker could inject a few malicious texts into the knowledge database of a RAG system to induce an LLM to generate an attacker-chosen target answer for an attacker-chosen target question. We formulate knowledge corruption attacks as an optimization problem, whose solution is a set of malicious texts. Depending on the background knowledge (e.g., black-box and white-box settings) of an attacker on a RAG system, we propose two solutions to solve the optimization problem, respectively. Our results show PoisonedRAG could achieve a 90% attack success rate when injecting five malicious texts for each target question into a knowledge database with millions of texts. We also evaluate several defenses and our results show they are insufficient to defend against PoisonedRAG, highlighting the need for new defenses.

Shuaijie She, Wei Zou, Shujian Huang et al.

Though reasoning abilities are considered language-agnostic, existing LLMs exhibit inconsistent reasoning abilities across different languages, e.g., reasoning in the dominant language like English is superior to other languages due to the imbalance of multilingual training data. To enhance reasoning abilities in non-dominant languages, we propose a Multilingual-Alignment-as-Preference Optimization framework (MAPO), aiming to align the reasoning processes in other languages with the dominant language. Specifically, we harness an off-the-shelf translation model for the consistency between answers in non-dominant and dominant languages, which we adopt as the preference for optimization, e.g., Direct Preference Optimization (DPO) or Proximal Policy Optimization (PPO). Experiments show that MAPO stably achieves significant improvements in the multilingual reasoning of various models on all three benchmarks (MSVAMP +16.2%, MGSM +6.1%, and MNumGLUESub +13.3%), with improved reasoning consistency across languages.

Na Liu, Liangyu Chen, Xiaoyu Tian et al.

This paper introduces RAISE (Reasoning and Acting through Scratchpad and Examples), an advanced architecture enhancing the integration of Large Language Models (LLMs) like GPT-4 into conversational agents. RAISE, an enhancement of the ReAct framework, incorporates a dual-component memory system, mirroring human short-term and long-term memory, to maintain context and continuity in conversations. It entails a comprehensive agent construction scenario, including phases like Conversation Selection, Scene Extraction, CoT Completion, and Scene Augmentation, leading to the LLMs Training phase. This approach appears to enhance agent controllability and adaptability in complex, multi-turn dialogues. Our preliminary evaluations in a real estate sales context suggest that RAISE has some advantages over traditional agents, indicating its potential for broader applications. This work contributes to the AI field by providing a robust framework for developing more context-aware and versatile conversational agents.

Cheng Wen, Tingwei Guo, Shuaijiang Zhao et al.

Recent work shows that reinforcement learning(RL) can markedly sharpen the reasoning ability of large language models (LLMs) by prompting them to "think before answering." Yet whether and how these gains transfer to audio-language reasoning remains largely unexplored. We extend the Group-Relative Policy Optimization (GRPO) framework from DeepSeek-R1 to a Large Audio-Language Model (LALM), and construct a 32k sample multiple-choice corpus. Using a two-stage regimen supervised fine-tuning on structured and unstructured chains-of-thought, followed by curriculum-guided GRPO, we systematically compare implicit vs. explicit, and structured vs. free form reasoning under identical architectures. Our structured audio reasoning model, SARI (Structured Audio Reasoning via Curriculum-Guided Reinforcement Learning), achieves a 16.35% improvement in average accuracy over the base model Qwen2-Audio-7B-Instruct. Furthermore, the variant built upon Qwen2.5-Omni reaches state-of-the-art performance of 67.08% on the MMAU test-mini benchmark. Ablation experiments show that on the base model we use: (i) SFT warm-up is important for stable RL training, (ii) structured chains yield more robust generalization than unstructured ones, and (iii) easy-to-hard curricula accelerate convergence and improve final performance. These findings demonstrate that explicit, structured reasoning and curriculum learning substantially enhances audio-language understanding.

Yuxin Guo, Shijie Ma, Hu Su et al.

Audio-Visual Source Localization (AVSL) aims to locate sounding objects within video frames given the paired audio clips. Existing methods predominantly rely on self-supervised contrastive learning of audio-visual correspondence. Without any bounding-box annotations, they struggle to achieve precise localization, especially for small objects, and suffer from blurry boundaries and false positives. Moreover, the naive semi-supervised method is poor in fully leveraging the information of abundant unlabeled data. In this paper, we propose a novel semi-supervised learning framework for AVSL, namely Dual Mean-Teacher (DMT), comprising two teacher-student structures to circumvent the confirmation bias issue. Specifically, two teachers, pre-trained on limited labeled data, are employed to filter out noisy samples via the consensus between their predictions, and then generate high-quality pseudo-labels by intersecting their confidence maps. The sufficient utilization of both labeled and unlabeled data and the proposed unbiased framework enable DMT to outperform current state-of-the-art methods by a large margin, with CIoU of 90.4% and 48.8% on Flickr-SoundNet and VGG-Sound Source, obtaining 8.9%, 9.6% and 4.6%, 6.4% improvements over self- and semi-supervised methods respectively, given only 3% positional-annotations. We also extend our framework to some existing AVSL methods and consistently boost their performance.

Yuxin Guo, Shuailei Ma, Shijie Ma et al.

Audio is essential for multimodal video understanding. On the one hand, video inherently contains audio, which supplies complementary information to vision. Besides, video large language models (Video-LLMs) can encounter many audio-centric settings. However, existing Video-LLMs and Audio-Visual Large Language Models (AV-LLMs) exhibit deficiencies in exploiting audio information, leading to weak understanding and hallucinations. To solve the issues, we delve into the model architecture and dataset. (1) From the architectural perspective, we propose a fine-grained AV-LLM, namely Dolphin. The concurrent alignment of audio and visual modalities in both temporal and spatial dimensions ensures a comprehensive and accurate understanding of videos. Specifically, we devise an audio-visual multi-scale adapter for multi-scale information aggregation, which achieves spatial alignment. For temporal alignment, we propose audio-visual interleaved merging. (2) From the dataset perspective, we curate an audio-visual caption and instruction-tuning dataset, called AVU. It comprises 5.2 million diverse, open-ended data tuples (video, audio, question, answer) and introduces a novel data partitioning strategy. Extensive experiments show our model not only achieves remarkable performance in audio-visual understanding, but also mitigates potential hallucinations.

Yuxin Guo, Shijie Ma, Yuhao Zhao et al.

Audio-Visual Source Localization (AVSL) is the task of identifying specific sounding objects in the scene given audio cues. In our work, we focus on semi-supervised AVSL with pseudo-labeling. To address the issues with vanilla hard pseudo-labels including bias accumulation, noise sensitivity, and instability, we propose a novel method named Cross Pseudo-Labeling (XPL), wherein two models learn from each other with the cross-refine mechanism to avoid bias accumulation. We equip XPL with two effective components. Firstly, the soft pseudo-labels with sharpening and pseudo-label exponential moving average mechanisms enable models to achieve gradual self-improvement and ensure stable training. Secondly, the curriculum data selection module adaptively selects pseudo-labels with high quality during training to mitigate potential bias. Experimental results demonstrate that XPL significantly outperforms existing methods, achieving state-of-the-art performance while effectively mitigating confirmation bias and ensuring training stability.

Yanting Wang, Hongye Fu, Wei Zou et al.

Different from a unimodal model whose input is from a single modality, the input (called multi-modal input) of a multi-modal model is from multiple modalities such as image, 3D points, audio, text, etc. Similar to unimodal models, many existing studies show that a multi-modal model is also vulnerable to adversarial perturbation, where an attacker could add small perturbation to all modalities of a multi-modal input such that the multi-modal model makes incorrect predictions for it. Existing certified defenses are mostly designed for unimodal models, which achieve sub-optimal certified robustness guarantees when extended to multi-modal models as shown in our experimental results. In our work, we propose MMCert, the first certified defense against adversarial attacks to a multi-modal model. We derive a lower bound on the performance of our MMCert under arbitrary adversarial attacks with bounded perturbations to both modalities (e.g., in the context of auto-driving, we bound the number of changed pixels in both RGB image and depth image). We evaluate our MMCert using two benchmark datasets: one for the multi-modal road segmentation task and the other for the multi-modal emotion recognition task. Moreover, we compare our MMCert with a state-of-the-art certified defense extended from unimodal models. Our experimental results show that our MMCert outperforms the baseline.

Wei Zou, Sen Yang, Yu Bao et al.

The rise of Large Language Models (LLMs) has reshaped machine translation (MT), but multilingual MT still relies heavily on parallel data for supervised fine-tuning (SFT), facing challenges like data scarcity for low-resource languages and catastrophic forgetting. To address these issues, we propose TRANS-ZERO, a self-play framework that leverages only monolingual data and the intrinsic multilingual knowledge of LLM. TRANS-ZERO combines Genetic Monte-Carlo Tree Search (G-MCTS) with preference optimization, achieving strong translation performance that rivals supervised methods. Experiments demonstrate that this approach not only matches the performance of models trained on large-scale parallel data but also excels in non-English translation directions. Further analysis reveals that G-MCTS itself significantly enhances translation quality by exploring semantically consistent candidates through iterative translations, providing a robust foundation for the framework's succuss.

Wei Zou, Ziyuan Zhuang, Xiang Geng et al.

Paraphrase generation strives to generate high-quality and diverse expressions of a given text, a domain where diffusion models excel. Though SOTA diffusion generation reconciles generation quality and diversity, textual diffusion suffers from a truncation issue that hinders efficiency and quality control. In this work, we propose \textit{L}atent \textit{D}iffusion \textit{P}araphraser~(LDP), a novel paraphrase generation by modeling a controllable diffusion process given a learned latent space. LDP achieves superior generation efficiency compared to its diffusion counterparts. It can facilitate only input segments to ensure paraphrase semantics, improving the results without external features. Experiments show that LDP better reconciles paraphrase generation quality and diversity than baselines. Further analysis shows that our method is also helpful to other similar text generations and domain adaptations

Wei Zou, Yupei Liu, Yanting Wang et al.

LLM-integrated applications are vulnerable to prompt injection attacks, where an attacker contaminates the input to inject malicious prompts, causing the LLM to follow the attacker's intent instead of the original user's. Existing prompt injection detection methods often have sub-optimal performance and/or high computational overhead. In this work, we propose PIShield, a detection method that is both effective and efficient. Our key observation is that the internal representation of the final token in a prompt-extracted from a specific layer of the LLM, which we term the injection-critical layer-captures distinguishing features between clean and contaminated prompts. Leveraging this insight, we train a simple linear classifier on these internal representations using a labeled set of clean and contaminated prompts. We compare PIShield against 11 baselines across 5 diverse benchmark datasets and 8 prompt injection attacks. The results demonstrate that PIShield is both highly effective and efficient, substantially outperforming existing methods. Additionally, we show that PIShield resists strong adaptive attacks.

Bajian Xiang, Shuaijiang Zhao, Tingwei Guo et al.

End-to-end Large Speech Language Models (LSLMs) have demonstrated impressive conversational generation abilities, yet consistently fall short of traditional pipeline systems on semantic understanding benchmarks. In this work, we reveal through systematic experimentation that although LSLMs lose some text input performance after speech-text alignment training, the performance gap between speech and text inputs is more pronounced, which we refer to as the modality gap. To understand this gap, we analyze both coarse- and fine-grained text and speech representations. At the coarse-grained level, representations of speech and text in deeper layers are found to be increasingly aligned in direction (cosine similarity), while concurrently diverging in magnitude (Euclidean distance). We further find that representation similarity is strongly correlated with the modality gap. At the fine-grained level, a spontaneous token-level alignment pattern between text and speech representations is observed. Based on this, we introduce the Alignment Path Score to quantify token-level alignment quality, which exhibits stronger correlation with the modality gap. Building on these insights, we design targeted interventions on critical tokens through angle projection and length normalization. These strategies demonstrate the potential to improve correctness for speech inputs. Our study provides the first systematic empirical analysis of the modality gap and alignment mechanisms in LSLMs, offering both theoretical and methodological guidance for future optimization.

Wei Zou, Shujian Huang, Jiajun Chen

Generating adversarial examples contributes to mainstream neural machine translation~(NMT) robustness. However, popular adversarial policies are apt for fixed tokenization, hindering its efficacy for common character perturbations involving versatile tokenization. Based on existing adversarial generation via reinforcement learning~(RL), we propose the `DexChar policy' that introduces character perturbations for the existing mainstream adversarial policy based on token substitution. Furthermore, we improve the self-supervised matching that provides feedback in RL to cater to the semantic constraints required during training adversaries. Experiments show that our method is compatible with the scenario where baseline adversaries fail, and can generate high-efficiency adversarial examples for analysis and optimization of the system.

Guoguo Chen, Shuzhou Chai, Guanbo Wang et al.

This paper introduces GigaSpeech, an evolving, multi-domain English speech recognition corpus with 10,000 hours of high quality labeled audio suitable for supervised training, and 40,000 hours of total audio suitable for semi-supervised and unsupervised training. Around 40,000 hours of transcribed audio is first collected from audiobooks, podcasts and YouTube, covering both read and spontaneous speaking styles, and a variety of topics, such as arts, science, sports, etc. A new forced alignment and segmentation pipeline is proposed to create sentence segments suitable for speech recognition training, and to filter out segments with low-quality transcription. For system training, GigaSpeech provides five subsets of different sizes, 10h, 250h, 1000h, 2500h, and 10000h. For our 10,000-hour XL training subset, we cap the word error rate at 4% during the filtering/validation stage, and for all our other smaller training subsets, we cap it at 0%. The DEV and TEST evaluation sets, on the other hand, are re-processed by professional human transcribers to ensure high transcription quality. Baseline systems are provided for popular speech recognition toolkits, namely Athena, ESPnet, Kaldi and Pika.

Dongsong Yu, Guangliang Yang, Guozhu Meng et al.

To investigate the status quo of SEAndroid policy customization, we propose SEPAL, a universal tool to automatically retrieve and examine the customized policy rules. SEPAL applies the NLP technique and employs and trains a wide&deep model to quickly and precisely predict whether one rule is unregulated or not.Our evaluation shows SEPAL is effective, practical and scalable. We verify SEPAL outperforms the state of the art approach (i.e., EASEAndroid) by 15% accuracy rate on average. In our experiments, SEPAL successfully identifies 7,111 unregulated policy rules with a low false positive rate from 595,236 customized rules (extracted from 774 Android firmware images of 72 manufacturers). We further discover the policy customization problem is getting worse in newer Android versions (e.g., around 8% for Android 7 and nearly 20% for Android 9), even though more and more efforts are made. Then, we conduct a deep study and discuss why the unregulated rules are introduced and how they can compromise user devices. Last, we report some unregulated rules to seven vendors and so far four of them confirm our findings.

Dongwei Jiang, Wubo Li, Miao Cao et al.

Self-supervised visual pretraining has shown significant progress recently. Among those methods, SimCLR greatly advanced the state of the art in self-supervised and semi-supervised learning on ImageNet. The input feature representations for speech and visual tasks are both continuous, so it is natural to consider applying similar objective on speech representation learning. In this paper, we propose Speech SimCLR, a new self-supervised objective for speech representation learning. During training, Speech SimCLR applies augmentation on raw speech and its spectrogram. Its objective is the combination of contrastive loss that maximizes agreement between differently augmented samples in the latent space and reconstruction loss of input representation. The proposed method achieved competitive results on speech emotion recognition and speech recognition.

Wubo Li, Dongwei Jiang, Wei Zou et al.

Audio Visual Scene-aware Dialog (AVSD) is a task to generate responses when discussing about a given video. The previous state-of-the-art model shows superior performance for this task using Transformer-based architecture. However, there remain some limitations in learning better representation of modalities. Inspired by Neural Machine Translation (NMT), we propose the Transformer-based Modal Translator (TMT) to learn the representations of the source modal sequence by translating the source modal sequence to the related target modal sequence in a supervised manner. Based on Multimodal Transformer Networks (MTN), we apply TMT to video and dialog, proposing MTN-TMT for the video-grounded dialog system. On the AVSD track of the Dialog System Technology Challenge 7, MTN-TMT outperforms the MTN and other submission models in both Video and Text task and Text Only task. Compared with MTN, MTN-TMT improves all metrics, especially, achieving relative improvement up to 14.1% on CIDEr. Index Terms: multimodal learning, audio-visual scene-aware dialog, neural machine translation, multi-task learning

Ruixiong Zhang, Haiwei Wu, Wubo Li et al.

Recently, a variety of acoustic tasks and related applications arised. For many acoustic tasks, the labeled data size may be limited. To handle this problem, we propose an unsupervised pre-training method using Transformer based encoder to learn a general and robust high-level representation for all acoustic tasks. Experiments have been conducted on three kinds of acoustic tasks: speech emotion recognition, sound event detection and speech translation. All the experiments have shown that pre-training using its own training data can significantly improve the performance. With a larger pre-training data combining MuST-C, Librispeech and ESC-US datasets, for speech emotion recognition, the UAR can further improve absolutely 4.3% on IEMOCAP dataset. For sound event detection, the F1 score can further improve absolutely 1.5% on DCASE2018 task5 development set and 2.1% on evaluation set. For speech translation, the BLEU score can further improve relatively 12.2% on En-De dataset and 8.4% on En-Fr dataset.

Dongwei Jiang, Wubo Li, Ruixiong Zhang et al.

Building a good speech recognition system usually requires large amounts of transcribed data, which is expensive to collect. To tackle this problem, many unsupervised pre-training methods have been proposed. Among these methods, Masked Predictive Coding achieved significant improvements on various speech recognition datasets with BERT-like Masked Reconstruction loss and Transformer backbone. However, many aspects of MPC have not been fully investigated. In this paper, we conduct a further study on MPC and focus on three important aspects: the effect of pre-training data speaking style, its extension on streaming model, and how to better transfer learned knowledge from pre-training stage to downstream tasks. Experiments reveled that pre-training data with matching speaking style is more useful on downstream recognition tasks. A unified training objective with APC and MPC provided 8.46% relative error reduction on streaming model trained on HKUST. Also, the combination of target data adaption and layer-wise discriminative training helped the knowledge transfer of MPC, which achieved 3.99% relative error reduction on AISHELL over a strong baseline.

Wei Zou, Shujian Huang, Jun Xie et al.

Neural machine translation systems tend to fail on less decent inputs despite its significant efficacy, which may significantly harm the credibility of this systems-fathoming how and when neural-based systems fail in such cases is critical for industrial maintenance. Instead of collecting and analyzing bad cases using limited handcrafted error features, here we investigate this issue by generating adversarial examples via a new paradigm based on reinforcement learning. Our paradigm could expose pitfalls for a given performance metric, e.g., BLEU, and could target any given neural machine translation architecture. We conduct experiments of adversarial attacks on two mainstream neural machine translation architectures, RNN-search, and Transformer. The results show that our method efficiently produces stable attacks with meaning-preserving adversarial examples. We also present a qualitative and quantitative analysis for the preference pattern of the attack, demonstrating its capability of pitfall exposure.

Wubo Li, Wei Zou, Xiangang Li

Multimodalities provide promising performance than unimodality in most tasks. However, learning the semantic of the representations from multimodalities efficiently is extremely challenging. To tackle this, we propose the Transformer based Cross-modal Translator (TCT) to learn unimodal sequence representations by translating from other related multimodal sequences on a supervised learning method. Combined TCT with Multimodal Transformer Network (MTN), we evaluate MTN-TCT on the video-grounded dialogue which uses multimodality. The proposed method reports new state-of-the-art performance on video-grounded dialogue which indicates representations learned by TCT are more semantics compared to directly use unimodality.

Ruixiong Zhang, Wei Zou, Xiangang Li

Acoustic scene classification (ASC) and acoustic event detection (AED) are different but related tasks. Acoustic events can provide useful information for recognizing acoustic scenes. However, most of the datasets are provided without either the acoustic event or scene labels. To utilize the acoustic event information to improve the performance of ASC tasks, we present the cross-task pre-training mechanism which utilizes acoustic event information from the pre-trained AED model for ASC tasks. On the other hand, most of the models were designed and implemented on platforms with rich computing resources, and the on-device applications were limited. To solve this problem, we use model distillation method to compress our cross-task model to enable on-device acoustic scene classification. In this paper, the cross-task models and their student model were trained and evaluated on two datasets: TAU Urban Acoustic Scenes 2019 dataset and TUT Acoustic Scenes 2017 dataset. Results have shown that cross-task pre-training mechanism can significantly improve the performance of ASC tasks. The performance of our best model improved relatively 9.5% in the TAU Urban Acoustic Scenes 2019 dataset, and also improved 10% in the TUT Acoustic Scenes 2017 dataset compared with the official baseline. At the same time, the performance of the student model is much better than that of the model without teachers.

Dongwei Jiang, Xiaoning Lei, Wubo Li et al.

Speech recognition technologies are gaining enormous popularity in various industrial applications. However, building a good speech recognition system usually requires large amounts of transcribed data, which is expensive to collect. To tackle this problem, an unsupervised pre-training method called Masked Predictive Coding is proposed, which can be applied for unsupervised pre-training with Transformer based model. Experiments on HKUST show that using the same training data, we can achieve CER 23.3%, exceeding the best end-to-end model by over 0.2% absolute CER. With more pre-training data, we can further reduce the CER to 21.0%, or a 11.8% relative CER reduction over baseline.

Hongxuan Ma, Wei Zou, Zheng Zhu et al.

In the field of navigation and visual servo, it is common to calculate relative pose by feature points on markers, so keeping markers in camera's view is an important problem. In this paper, we propose a novel approach to calculate field-of-view (FOV) constraint of markers for camera. Our method can make the camera maintain the visibility of all feature points during the motion of mobile robot. According to the angular aperture of camera, the mobile robot can obtain the FOV constraint region where the camera cannot keep all feature points in an image. Based on the FOV constraint region, the mobile robot can be guided to move from the initial position to destination. Finally simulations and experiments are conducted based on a mobile robot equipped with a pan-tilt camera, which validates the effectiveness of the method to obtain the FOV constraints.

Zheng Zhu, Hongxuan Ma, Wei Zou

Human following on mobile robots has witnessed significant advances due to its potentials for real-world applications. Currently most human following systems are equipped with depth sensors to obtain distance information between human and robot, which suffer from the perception requirements and noises. In this paper, we design a wheeled mobile robot system with monocular pan-tilt camera to follow human, which can stay the target in the field of view and keep following simultaneously. The system consists of fast human detector, real-time and accurate visual tracker, and unified controller for mobile robot and pan-tilt camera. In visual tracking algorithm, both Siamese networks and optical flow information are exploited to locate and regress human simultaneously. In order in perform following with a monocular camera, the constraint of human height is introduced to design the controller. In experiments, human following are conducted and analysed in simulations and a real robot platform, which demonstrate the effectiveness and robustness of the overall system.

Zheng Zhu, Wei Zou, Guan Huang et al.

Convolutional neural networks (CNN) based tracking approaches have shown favorable performance in recent benchmarks. Nonetheless, the chosen CNN features are always pre-trained in different tasks and individual components in tracking systems are learned separately, thus the achieved tracking performance may be suboptimal. Besides, most of these trackers are not designed towards real-time applications because of their time-consuming feature extraction and complex optimization details. In this paper, we propose an end-to-end framework to learn the convolutional features and perform the tracking process simultaneously, namely, a unified convolutional tracker (UCT). Specifically, the UCT treats feature extractor and tracking process both as convolution operation and trains them jointly, which enables learned CNN features are tightly coupled with tracking process. During online tracking, an efficient model updating method is proposed by introducing peak-versus-noise ratio (PNR) criterion, and scale changes are handled efficiently by incorporating a scale branch into network. Experiments are performed on four challenging tracking datasets: OTB2013, OTB2015, VOT2015 and VOT2016. Our method achieves leading performance on these benchmarks while maintaining beyond real-time speed.

Zhaobing Kang, Wei Zou, Zheng Zhu

Accurate camera pose estimation result is essential for visual SLAM (VSLAM). This paper presents a novel pose correction method to improve the accuracy of the VSLAM system. Firstly, the relationship between the camera pose estimation error and bias values of map points is derived based on the optimized function in VSLAM. Secondly, the bias value of the map point is calculated by a statistical method. Finally, the camera pose estimation error is compensated according to the first derived relationship. After the pose correction, procedures of the original system, such as the bundle adjustment (BA) optimization, can be executed as before. Compared with existing methods, our algorithm is compact and effective and can be easily generalized to different VSLAM systems. Additionally, the robustness to system noise of our method is better than feature selection methods, due to all original system information is preserved in our algorithm while only a subset is employed in the latter. Experimental results on benchmark datasets show that our approach leads to considerable improvements over state-of-the-art algorithms for absolute pose estimation.

Jiabin Zhang, Zheng Zhu, Wei Zou et al.

Both accuracy and efficiency are significant for pose estimation and tracking in videos. State-of-the-art performance is dominated by two-stages top-down methods. Despite the leading results, these methods are impractical for real-world applications due to their separated architectures and complicated calculation. This paper addresses the task of articulated multi-person pose estimation and tracking towards real-time speed. An end-to-end multi-task network (MTN) is designed to perform human detection, pose estimation, and person re-identification (Re-ID) tasks simultaneously. To alleviate the performance bottleneck caused by scale variation problem, a paradigm which exploits scale-normalized image and feature pyramids (SIFP) is proposed to boost both performance and speed. Given the results of MTN, we adopt an occlusion-aware Re-ID feature strategy in the pose tracking module, where pose information is utilized to infer the occlusion state to make better use of Re-ID feature. In experiments, we demonstrate that the pose estimation and tracking performance improves steadily utilizing SIFP through different backbones. Using ResNet-18 and ResNet-50 as backbones, the overall pose tracking framework achieves competitive performance with 29.4 FPS and 12.2 FPS, respectively. Additionally, occlusion-aware Re-ID feature decreases the identification switches by 37% in the pose tracking process.

Kun Han, Junwen Chen, Hui Zhang et al.

In this paper we present DELTA, a deep learning based language technology platform. DELTA is an end-to-end platform designed to solve industry level natural language and speech processing problems. It integrates most popular neural network models for training as well as comprehensive deployment tools for production. DELTA aims to provide easy and fast experiences for using, deploying, and developing natural language processing and speech models for both academia and industry use cases. We demonstrate the reliable performance with DELTA on several natural language processing and speech tasks, including text classification, named entity recognition, natural language inference, speech recognition, speaker verification, etc. DELTA has been used for developing several state-of-the-art algorithms for publications and delivering real production to serve millions of users.