Jiaqing Liu

Shurong Chai, Rahul Kumar Jain, Shiyu Teng et al.

Recently, foundation models have been introduced demonstrating various tasks in the field of computer vision. These models such as Segment Anything Model (SAM) are generalized models trained using huge datasets. Currently, ongoing research focuses on exploring the effective utilization of these generalized models for specific domains, such as medical imaging. However, in medical imaging, the lack of training samples due to privacy concerns and other factors presents a major challenge for applying these generalized models to medical image segmentation task. To address this issue, the effective fine tuning of these models is crucial to ensure their optimal utilization. In this study, we propose to combine a complementary Convolutional Neural Network (CNN) along with the standard SAM network for medical image segmentation. To reduce the burden of fine tuning large foundation model and implement cost-efficient trainnig scheme, we focus only on fine-tuning the additional CNN network and SAM decoder part. This strategy significantly reduces trainnig time and achieves competitive results on publicly available dataset. The code is available at https://github.com/11yxk/SAM-LST.

Hao Sun, Hongyi Wang, Jiaqing Liu et al.

Multimodal sentiment analysis and depression estimation are two important research topics that aim to predict human mental states using multimodal data. Previous research has focused on developing effective fusion strategies for exchanging and integrating mind-related information from different modalities. Some MLP-based techniques have recently achieved considerable success in a variety of computer vision tasks. Inspired by this, we explore multimodal approaches with a feature-mixing perspective in this study. To this end, we introduce CubeMLP, a multimodal feature processing framework based entirely on MLP. CubeMLP consists of three independent MLP units, each of which has two affine transformations. CubeMLP accepts all relevant modality features as input and mixes them across three axes. After extracting the characteristics using CubeMLP, the mixed multimodal features are flattened for task predictions. Our experiments are conducted on sentiment analysis datasets: CMU-MOSI and CMU-MOSEI, and depression estimation dataset: AVEC2019. The results show that CubeMLP can achieve state-of-the-art performance with a much lower computing cost.

Qian Chen, Wen Wang, Qinglin Zhang et al.

Recently, unified speech-text models, such as SpeechGPT, VioLA, and AudioPaLM, have achieved remarkable performance on various speech tasks. These models discretize speech signals into tokens (speech discretization) and use a shared vocabulary for both text and speech tokens. Then they train a single decoder-only Transformer on a mixture of speech tasks. However, these models rely on the Loss Masking strategy for the ASR task, which ignores the dependency among speech tokens. In this paper, we propose to model speech tokens in an autoregressive way, similar to text. We find that applying the conventional cross-entropy loss on input speech tokens does not consistently improve the ASR performance over the Loss Masking approach. To address this issue, we propose a novel approach denoted Smoothed Label Distillation (SLD), which applies a KL divergence loss with smoothed labels on speech tokens. Our experiments show that SLD effectively models speech tokens and outperforms Loss Masking for decoder-only Transformers in ASR tasks with different speech discretization methods. The source code can be found here: https://github.com/alibaba-damo-academy/SpokenNLP/tree/main/sld

Qinglin Zhang, Chong Deng, Jiaqing Liu et al.

Listening to long video/audio recordings from video conferencing and online courses for acquiring information is extremely inefficient. Even after ASR systems transcribe recordings into long-form spoken language documents, reading ASR transcripts only partly speeds up seeking information. It has been observed that a range of NLP applications, such as keyphrase extraction, topic segmentation, and summarization, significantly improve users' efficiency in grasping important information. The meeting scenario is among the most valuable scenarios for deploying these spoken language processing (SLP) capabilities. However, the lack of large-scale public meeting datasets annotated for these SLP tasks severely hinders their advancement. To prompt SLP advancement, we establish a large-scale general Meeting Understanding and Generation Benchmark (MUG) to benchmark the performance of a wide range of SLP tasks, including topic segmentation, topic-level and session-level extractive summarization and topic title generation, keyphrase extraction, and action item detection. To facilitate the MUG benchmark, we construct and release a large-scale meeting dataset for comprehensive long-form SLP development, the AliMeeting4MUG Corpus, which consists of 654 recorded Mandarin meeting sessions with diverse topic coverage, with manual annotations for SLP tasks on manual transcripts of meeting recordings. To the best of our knowledge, the AliMeeting4MUG Corpus is so far the largest meeting corpus in scale and facilitates most SLP tasks. In this paper, we provide a detailed introduction of this corpus, SLP tasks and evaluation methods, baseline systems and their performance.

Qian Chen, Luyao Cheng, Chong Deng et al.

Recent advancements in joint speech-text models show great potential for seamless voice interactions. However, existing models face critical challenges: temporal resolution mismatch between speech tokens (25Hz) and text tokens (~3Hz) dilutes semantic information, incurs high computational costs, and causes catastrophic forgetting of text LLM knowledge. We introduce Fun-Audio-Chat, a Large Audio Language Model addressing these limitations via two innovations from our previous work DrVoice. First, Dual-Resolution Speech Representations (DRSR): the Shared LLM processes audio at efficient 5Hz (via token grouping), while the Speech Refined Head generates high-quality tokens at 25Hz, balancing efficiency (~50% GPU reduction) and quality. Second, Core-Cocktail Training, a two-stage fine-tuning with intermediate merging that mitigates catastrophic forgetting. We then apply Multi-Task DPO Training to enhance robustness, audio understanding, instruction-following and voice empathy. This multi-stage post-training enables Fun-Audio-Chat to retain text LLM knowledge while gaining powerful audio understanding, reasoning, and generation. Unlike recent LALMs requiring large-scale audio-text pre-training, Fun-Audio-Chat leverages pre-trained models and extensive post-training. Fun-Audio-Chat 8B and MoE 30B-A3B achieve competitive performance on Speech-to-Text and Speech-to-Speech tasks, ranking top among similar-scale models on Spoken QA benchmarks. They also achieve competitive to superior performance on Audio Understanding, Speech Function Calling, Instruction-Following and Voice Empathy. We develop Fun-Audio-Chat-Duplex, a full-duplex variant with strong performance on Spoken QA and full-duplex interactions. We open-source Fun-Audio-Chat-8B with training and inference code, and provide an interactive demo.

Hai Yu, Chong Deng, Qinglin Zhang et al.

Topic segmentation is critical for obtaining structured documents and improving downstream tasks such as information retrieval. Due to its ability of automatically exploring clues of topic shift from abundant labeled data, recent supervised neural models have greatly promoted the development of long document topic segmentation, but leaving the deeper relationship between coherence and topic segmentation underexplored. Therefore, this paper enhances the ability of supervised models to capture coherence from both logical structure and semantic similarity perspectives to further improve the topic segmentation performance, proposing Topic-aware Sentence Structure Prediction (TSSP) and Contrastive Semantic Similarity Learning (CSSL). Specifically, the TSSP task is proposed to force the model to comprehend structural information by learning the original relations between adjacent sentences in a disarrayed document, which is constructed by jointly disrupting the original document at topic and sentence levels. Moreover, we utilize inter- and intra-topic information to construct contrastive samples and design the CSSL objective to ensure that the sentences representations in the same topic have higher similarity, while those in different topics are less similar. Extensive experiments show that the Longformer with our approach significantly outperforms old state-of-the-art (SOTA) methods. Our approach improve $F_1$ of old SOTA by 3.42 (73.74 -> 77.16) and reduces $P_k$ by 1.11 points (15.0 -> 13.89) on WIKI-727K and achieves an average relative reduction of 4.3% on $P_k$ on WikiSection. The average relative $P_k$ drop of 8.38% on two out-of-domain datasets also demonstrates the robustness of our approach.

Qinglin Zhang, Chong Deng, Jiaqing Liu et al.

ICASSP2023 General Meeting Understanding and Generation Challenge (MUG) focuses on prompting a wide range of spoken language processing (SLP) research on meeting transcripts, as SLP applications are critical to improve users' efficiency in grasping important information in meetings. MUG includes five tracks, including topic segmentation, topic-level and session-level extractive summarization, topic title generation, keyphrase extraction, and action item detection. To facilitate MUG, we construct and release a large-scale meeting dataset, the AliMeeting4MUG Corpus.

Shiyu Teng, Jiaqing Liu, Hao Sun et al.

Depression remains widely underdiagnosed and undertreated because stigma and subjective symptom ratings hinder reliable screening. To address this challenge, we propose a coarse-to-fine, multi-stage framework that leverages large language models (LLMs) for accurate and interpretable detection. The pipeline performs binary screening, five-class severity classification, and continuous regression. At each stage, an LLM produces progressively richer clinical summaries that guide a multimodal fusion module integrating text, audio, and video features, yielding predictions with transparent rationale. The system then consolidates all summaries into a concise, human-readable assessment report. Experiments on the E-DAIC and CMDC datasets show significant improvements over state-of-the-art baselines in both accuracy and interpretability.

Hao-Xiang Xu, Chong Deng, Jiaqing Liu et al.

Large Language Models (LLMs) extend their capabilities through function-calling (FC), which relies on training data with high quality, diversity, and broad coverage of scenario. However, obtaining and annotating real function-calling data is challenging, while synthetic data from existing pipelines often suffers from unreliable APIs, limited tool scalability, insufficient diversity, and weak quality control. To address these, we present GenesisFunc, an automated pipeline for generating FC training data. Starting from reliable tools in widely used public benchmarks, our GenesisFunc employs a multi-agent framework to support a dialogue generation system that produces conversations spanning diverse scenarios, while maintaining both diversity and quality throughout the process. The accuracy of the data is further reinforced through a multi-stage evaluation system. We fine-tune an 8B LLM on the synthetic dataset and show through extensive experiments that it outperforms similarly sized open-source models in in-domain FC performance and out-of-domain generalization, while reaching FC capabilities comparable to some of the latest API-based models. In addition, our method demonstrates strong potential to scale effectively across downstream tools, underscoring its real-world applicability.

Jiaqing Liu, Chong Deng, Qinglin Zhang et al.

Meetings are increasingly important for collaborations. Action items in meeting transcripts are crucial for managing post-meeting to-do tasks, which usually are summarized laboriously. The Action Item Detection task aims to automatically detect meeting content associated with action items. However, datasets manually annotated with action item detection labels are scarce and in small scale. We construct and release the first Chinese meeting corpus with manual action item annotations. In addition, we propose a Context-Drop approach to utilize both local and global contexts by contrastive learning, and achieve better accuracy and robustness for action item detection. We also propose a Lightweight Model Ensemble method to exploit different pre-trained models. Experimental results on our Chinese meeting corpus and the English AMI corpus demonstrate the effectiveness of the proposed approaches.

Jiaqing Liu, Chong Deng, Qinglin Zhang et al.

Automatic Speech Recognition (ASR) transcripts exhibit recognition errors and various spoken language phenomena such as disfluencies, ungrammatical sentences, and incomplete sentences, hence suffering from poor readability. To improve readability, we propose a Contextualized Spoken-to-Written conversion (CoS2W) task to address ASR and grammar errors and also transfer the informal text into the formal style with content preserved, utilizing contexts and auxiliary information. This task naturally matches the in-context learning capabilities of Large Language Models (LLMs). To facilitate comprehensive comparisons of various LLMs, we construct a document-level Spoken-to-Written conversion of ASR Transcripts Benchmark (SWAB) dataset. Using SWAB, we study the impact of different granularity levels on the CoS2W performance, and propose methods to exploit contexts and auxiliary information to enhance the outputs. Experimental results reveal that LLMs have the potential to excel in the CoS2W task, particularly in grammaticality and formality, our methods achieve effective understanding of contexts and auxiliary information by LLMs. We further investigate the effectiveness of using LLMs as evaluators and find that LLM evaluators show strong correlations with human evaluations on rankings of faithfulness and formality, which validates the reliability of LLM evaluators for the CoS2W task.

Hao Sun, Yu Song, Jiaqing Liu et al.

Large-scale models have exhibited remarkable capabilities across diverse domains, including automated medical services and intelligent customer support. However, as most large models are trained on single-modality corpora, enabling them to effectively process and understand multimodal signals remains a significant challenge. Current research often focuses on designing task-specific or scenario-specific tuning strategies, which limits the scalability and versatility. To address this limitation, we propose a unified framework that concurrently handles multiple tasks and modalities. In this framework, all modalities and tasks are represented as unified tokens and trained using a single, consistent approach. To enable efficient multitask processing, we introduce a novel tuning strategy termed neural tuning, inspired by the concept of sparse distributed representation in the human brain, where only specific subsets of neurons are activated for each task. Furthermore, to advance research in multimodal and multitask learning, we present a new benchmark, MMUD, which includes samples annotated with multiple task labels spanning reasoning segmentation, referring segmentation, image captioning, and text-to-image generation. By applying neural tuning to pretrained large models on the MMUD benchmark, we demonstrate the ability to handle multiple tasks simultaneously in a streamlined and efficient manner. All models, code, and datasets will be released publicly upon publication, fostering further research and innovation in this field.

Hai Yu, Chong Deng, Qinglin Zhang et al.

The video topic segmentation (VTS) task segments videos into intelligible, non-overlapping topics, facilitating efficient comprehension of video content and quick access to specific content. VTS is also critical to various downstream video understanding tasks. Traditional VTS methods using shallow features or unsupervised approaches struggle to accurately discern the nuances of topical transitions. Recently, supervised approaches have achieved superior performance on video action or scene segmentation over unsupervised approaches. In this work, we improve supervised VTS by thoroughly exploring multimodal fusion and multimodal coherence modeling. Specifically, (1) we enhance multimodal fusion by exploring different architectures using cross-attention and mixture of experts. (2) To generally strengthen multimodality alignment and fusion, we pre-train and fine-tune the model with multimodal contrastive learning. (3) We propose a new pre-training task tailored for the VTS task, and a novel fine-tuning task for enhancing multimodal coherence modeling for VTS. We evaluate the proposed approaches on educational videos, in the form of lectures, due to the vital role of topic segmentation of educational videos in boosting learning experiences. Additionally, we introduce a large-scale Chinese lecture video dataset to augment the existing English corpus, promoting further research in VTS. Experiments on both English and Chinese lecture datasets demonstrate that our model achieves superior VTS performance compared to competitive unsupervised and supervised baselines.

Shurong Chai, Rahul Kumar JAIN, Rui Xu et al.

Deep learning relies heavily on data augmentation to mitigate limited data, especially in medical imaging. Recent multimodal learning integrates text and images for segmentation, known as referring or text-guided image segmentation. However, common augmentations like rotation and flipping disrupt spatial alignment between image and text, weakening performance. To address this, we propose an early fusion framework that combines text and visual features before augmentation, preserving spatial consistency. We also design a lightweight generator that projects text embeddings into visual space, bridging semantic gaps. Visualization of generated pseudo-images shows accurate region localization. Our method is evaluated on three medical imaging tasks and four segmentation frameworks, achieving state-of-the-art results. Code is publicly available on GitHub: https://github.com/11yxk/MedSeg_EarlyFusion.

Chao-Hong Tan, Qian Chen, Wen Wang et al.

Recent studies on end-to-end (E2E) speech generation with large language models (LLMs) have attracted significant community attention, with multiple works extending text-based LLMs to generate discrete speech tokens. Existing E2E approaches primarily fall into two categories: (1) Methods that generate discrete speech tokens independently without incorporating them into the LLM's autoregressive process, resulting in text generation being unaware of concurrent speech synthesis. (2) Models that generate interleaved or parallel speech-text tokens through joint autoregressive modeling, enabling mutual modality awareness during generation. This paper presents DrVoice, a parallel speech-text voice conversation model based on joint autoregressive modeling, featuring dual-resolution speech representations. Notably, while current methods utilize mainly 12.5Hz input audio representation, our proposed dual-resolution mechanism reduces the input frequency for the LLM to 5Hz, significantly reducing computational cost and alleviating the frequency discrepancy between speech and text tokens and in turn better exploiting LLMs' capabilities. Experimental results demonstrate that DRVOICE-7B establishes new state-of-the-art (SOTA) on OpenAudioBench and Big Bench Audio benchmarks, while achieving performance comparable to the SOTA on VoiceBench and UltraEval-Audio benchmarks, making it a leading open-source speech foundation model in ~7B models.

Qian Chen, Yafeng Chen, Yanni Chen et al.

Recent advancements in large language models (LLMs) and multimodal speech-text models have laid the groundwork for seamless voice interactions, enabling real-time, natural, and human-like conversations. Previous models for voice interactions are categorized as native and aligned. Native models integrate speech and text processing in one framework but struggle with issues like differing sequence lengths and insufficient pre-training. Aligned models maintain text LLM capabilities but are often limited by small datasets and a narrow focus on speech tasks. In this work, we introduce MinMo, a Multimodal Large Language Model with approximately 8B parameters for seamless voice interaction. We address the main limitations of prior aligned multimodal models. We train MinMo through multiple stages of speech-to-text alignment, text-to-speech alignment, speech-to-speech alignment, and duplex interaction alignment, on 1.4 million hours of diverse speech data and a broad range of speech tasks. After the multi-stage training, MinMo achieves state-of-the-art performance across various benchmarks for voice comprehension and generation while maintaining the capabilities of text LLMs, and also facilitates full-duplex conversation, that is, simultaneous two-way communication between the user and the system. Moreover, we propose a novel and simple voice decoder that outperforms prior models in voice generation. The enhanced instruction-following capabilities of MinMo supports controlling speech generation based on user instructions, with various nuances including emotions, dialects, and speaking rates, and mimicking specific voices. For MinMo, the speech-to-text latency is approximately 100ms, full-duplex latency is approximately 600ms in theory and 800ms in practice. The MinMo project web page is https://funaudiollm.github.io/minmo, and the code and models will be released soon.

Qinglin Zhang, Luyao Cheng, Chong Deng et al.

Full-duplex spoken dialogue systems significantly surpass traditional turn-based dialogue systems, as they allow simultaneous bidirectional communication, closely mirroring human-human interactions. However, achieving low latency and natural interactions in full-duplex dialogue systems remains a significant challenge, especially considering human conversation dynamics such as interruptions, backchannels, and overlapping speech. In this paper, we introduce a novel End-to-End GPT-based model OmniFlatten for full-duplex conversation, capable of effectively modeling the complex behaviors inherent to natural conversations with low latency. To achieve full-duplex conversation capabilities, we propose a multi-stage post-training scheme that progressively adapts a text large language model (LLM) backbone into a speech-text dialogue LLM, capable of generating text and speech in real time, without modifying the architecture of the backbone LLM. The training process comprises three stages: modality alignment, half-duplex dialogue learning, and full-duplex dialogue learning. In all training stages, we standardize the data using a flattening operation, which enables unifying the training methods and the GPT backbone across different modalities and tasks. Our approach offers a simple modeling technique and a promising research direction for developing efficient and natural end-to-end full-duplex spoken dialogue systems. Audio samples of dialogues generated by OmniFlatten can be found at this web site (https://omniflatten.github.io/).

Shiyu Teng, Jiaqing Liu, Rahul Kumar Jain et al.

Depression is one of the leading causes of disability worldwide, posing a severe burden on individuals, healthcare systems, and society at large. Recent advancements in Large Language Models (LLMs) have shown promise in addressing mental health challenges, including the detection of depression through text-based analysis. However, current LLM-based methods often struggle with nuanced symptom identification and lack a transparent, step-by-step reasoning process, making it difficult to accurately classify and explain mental health conditions. To address these challenges, we propose a Chain-of-Thought Prompting approach that enhances both the performance and interpretability of LLM-based depression detection. Our method breaks down the detection process into four stages: (1) sentiment analysis, (2) binary depression classification, (3) identification of underlying causes, and (4) assessment of severity. By guiding the model through these structured reasoning steps, we improve interpretability and reduce the risk of overlooking subtle clinical indicators. We validate our method on the E-DAIC dataset, where we test multiple state-of-the-art large language models. Experimental results indicate that our Chain-of-Thought Prompting technique yields superior performance in both classification accuracy and the granularity of diagnostic insights, compared to baseline approaches.

Zhenyu Liao, Jiaqing Liu, TianQi Hou et al.

Attention mechanisms have revolutionized machine learning (ML) by enabling efficient modeling of global dependencies across inputs. Their inherently parallelizable structures allow for efficient scaling with the exponentially increasing size of both pretrained data and model parameters. Yet, despite their central role as the computational backbone of modern large language models (LLMs), the theoretical understanding of Attentions, especially in the nonlinear setting, remains limited. In this paper, we provide a precise characterization of the \emph{in-context memorization error} of \emph{nonlinear Attention}, in the high-dimensional proportional regime where the number of input tokens $n$ and their embedding dimension $p$ are both large and comparable. Leveraging recent advances in the theory of large kernel random matrices, we show that nonlinear Attention typically incurs higher memorization error than linear ridge regression on random inputs. However, this gap vanishes, and can even be reversed, when the input exhibits statistical structure, particularly when the Attention weights align with the input signal direction. Our results reveal how nonlinearity and input structure interact with each other to govern the memorization performance of nonlinear Attention. The theoretical insights are supported by numerical experiments.

Ruibo Hou, Shiyu Teng, Jiaqing Liu et al.

Multimodal deep learning has shown promise in depression detection by integrating text, audio, and video signals. Recent work leverages sentiment analysis to enhance emotional understanding, yet suffers from high computational cost, domain mismatch, and static knowledge limitations. To address these issues, we propose a novel Retrieval-Augmented Generation (RAG) framework. Given a depression-related text, our method retrieves semantically relevant emotional content from a sentiment dataset and uses a Large Language Model (LLM) to generate an Emotion Prompt as an auxiliary modality. This prompt enriches emotional representation and improves interpretability. Experiments on the AVEC 2019 dataset show our approach achieves state-of-the-art performance with CCC of 0.593 and MAE of 3.95, surpassing previous transfer learning and multi-task learning baselines.

Qian Chen, Wen Wang, Qinglin Zhang et al.

The Transformer architecture has significantly advanced deep learning, particularly in natural language processing, by effectively managing long-range dependencies. However, as the demand for understanding complex relationships grows, refining the Transformer's architecture becomes critical. This paper introduces Skip-Layer Attention (SLA) to enhance Transformer models by enabling direct attention between non-adjacent layers. This method improves the model's ability to capture dependencies between high-level abstract features and low-level details. By facilitating direct attention between these diverse feature levels, our approach overcomes the limitations of current Transformers, which often rely on suboptimal intra-layer attention. Our implementation extends the Transformer's functionality by enabling queries in a given layer to interact with keys and values from both the current layer and one preceding layer, thus enhancing the diversity of multi-head attention without additional computational burden. Extensive experiments demonstrate that our enhanced Transformer model achieves superior performance in language modeling tasks, highlighting the effectiveness of our skip-layer attention mechanism.

Qian Chen, Wen Wang, Qinglin Zhang et al.

Prior studies diagnose the anisotropy problem in sentence representations from pre-trained language models, e.g., BERT, without fine-tuning. Our analysis reveals that the sentence embeddings from BERT suffer from a bias towards uninformative words, limiting the performance in semantic textual similarity (STS) tasks. To address this bias, we propose a simple and efficient unsupervised approach, Diagonal Attention Pooling (Ditto), which weights words with model-based importance estimations and computes the weighted average of word representations from pre-trained models as sentence embeddings. Ditto can be easily applied to any pre-trained language model as a postprocessing operation. Compared to prior sentence embedding approaches, Ditto does not add parameters nor requires any learning. Empirical evaluations demonstrate that our proposed Ditto can alleviate the anisotropy problem and improve various pre-trained models on STS tasks.

Qinglin Zhang, Qian Chen, Yali Li et al.

Transcripts generated by automatic speech recognition (ASR) systems for spoken documents lack structural annotations such as paragraphs, significantly reducing their readability. Automatically predicting paragraph segmentation for spoken documents may both improve readability and downstream NLP performance such as summarization and machine reading comprehension. We propose a sequence model with self-adaptive sliding window for accurate and efficient paragraph segmentation. We also propose an approach to exploit phonetic information, which significantly improves robustness of spoken document segmentation to ASR errors. Evaluations are conducted on the English Wiki-727K document segmentation benchmark, a Chinese Wikipedia-based document segmentation dataset we created, and an in-house Chinese spoken document dataset. Our proposed model outperforms the state-of-the-art (SOTA) model based on the same BERT-Base, increasing segmentation F1 on the English benchmark by 4.2 points and on Chinese datasets by 4.3-10.1 points, while reducing inference time to less than 1/6 of inference time of the current SOTA.