While fusing the capacities and advantages of various large language models offers a pathway to construct more powerful and versatile models, a fundamental challenge is to properly select advantageous model during training. Existing fusion methods primarily focus on the training mode that uses cross entropy on ground truth in a teacher-forcing setup to measure a model's advantage, which may provide limited insight towards model advantage. In this paper, we introduce a novel approach that enhances the fusion process by incorporating both the training and inference modes. Our method evaluates model advantage not only through cross entropy during training but also by considering inference outputs, providing a more comprehensive assessment. To combine the two modes effectively, we introduce ProFuser to progressively transition from inference mode to training mode. To validate ProFuser's effectiveness, we fused three models, including Vicuna-7B-v1.5, Llama-2-7B-Chat, and MPT-7B-8K-Chat, and demonstrated the improved performance in knowledge, reasoning, and safety compared to baseline methods.

Fuwen Luo, Shengfeng Lou, Chi Chen, Ziyue Wang, Chenliang Li, Weizhou Shen, Jiyue Guo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Yang Liu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Anwen Hu, Haiyang Xu, Liang Zhang, Jiabo Ye, Ming Yan, Ji Zhang, Qin Jin, Fei Huang, Jingren Zhou. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Ziheng Qiao, Houquan Zhou, Yumeng Liu, Zhenghua Li, Min Zhang, Bo Zhang, Chen Li, Ji Zhang, Fei Huang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Recently, model merging methods have demonstrated powerful strengths in combining abilities on various tasks from multiple Large Language Models (LLMs). While previous model merging methods mainly focus on merging homogeneous models with identical architecture, they meet challenges when dealing with Multimodal Large Language Models (MLLMs) with inherent heterogeneous property, including differences in model architecture and the asymmetry in the parameter space. In this work, we propose AdaMMS1, a novel model merging method tailored for heterogeneous MLLMs. Our method tackles the challenges in three steps: mapping, merging and searching. Specifically, we first design mapping function between models to apply model merging on MLLMs with different architecture. Then we apply linear interpolation on model weights to actively adapt the asymmetry in the heterogeneous MLLMs. Finally in the hyper-parameter searching step, we propose an unsupervised hyper-parameter selection method for model merging. As the first model merging method capable of merging heterogeneous MLLMs without labeled data, extensive experiments on various model combinations demonstrated that AdaMMS outperforms previous model merging methods on various vision-language benchmarks.2

As language models continue to scale, Large Language Models (LLMs) have exhibited emerging capabilities in In-Context Learning (ICL), enabling them to solve language tasks by prefixing a few in-context demonstrations (ICDs) as context. Inspired by these advancements, researchers have extended these techniques to develop Large Multimodal Models (LMMs) with ICL capabilities. However, existing LMMs face a critical issue: they often fail to effectively leverage the visual context in multimodal demonstrations and instead simply follow textual patterns. This indicates that LMMs do not achieve effective alignment between multimodal demonstrations and model outputs. To address this problem, we propose Symbol Demonstration Direct Preference Optimization (SymDPO). Specifically, SymDPO aims to break the traditional paradigm of constructing multimodal demonstrations by using random symbols to replace text answers within instances. This forces the model to carefully understand the demonstration images and establish a relationship between the images and the symbols to answer questions correctly. We validate the effectiveness of this method on multiple benchmarks, demonstrating that with SymDPO, LMMs can more effectively understand the multimodal context within examples and utilize this knowledge to answer questions better. Code is available at https://github.com/APiaoG/SymDPO.

Recent methods that integrate spatial layouts with text for document understanding in large language models (LLMs) have shown promising results. A commonly used method is to represent layout information as text tokens and interleave them with text content as inputs to the LLMs. However, such a method still demonstrates limitations, as it requires additional position IDs for tokens that are used to represent layout information. Due to the constraint on max position IDs, assigning them to layout information reduces those available for text content, reducing the capacity for the model to learn from the text during training, while also introducing a large number of potentially untrained position IDs during long-context inference, which can hinder performance on document understanding tasks. To address these issues, we propose LayTokenLLM, a simple yet effective method for document understanding. LayTokenLLM represents layout information as a single token per text segment and uses a specialized positional encoding scheme. It shares position IDs between text and layout tokens, eliminating the need for additional position IDs. This design maintains the model’s capacity to learn from text while mitigating long-context issues during inference. Furthermore, a novel pre-training objective called Next Interleaved Text and Layout Token Prediction (NTLP) is devised to enhance cross-modality learning between text and layout tokens. Extensive experiments show that LayTokenLLM outperforms existing layout-integrated LLMs and MLLMs of similar scales on multi-page document understanding tasks, as well as most single-page tasks.

Multi-modal Large Language Models have demonstrated remarkable capabilities in executing instructions for a variety of single-image tasks. Despite this progress, significant challenges remain in modeling long image sequences. In this work, we introduce the versatile multi-modal large language model, mPLUG-Owl3, which enhances the capability for long image-sequence understanding in scenarios that incorporate retrieved image-text knowledge, multimodal in-context examples, and lengthy videos. Specifically, we propose novel hyper attention blocks to efficiently integrate vision and language into a common language-guided semantic space, thereby facilitating the processing of extended multi-image scenarios. We conduct evaluations on 21 benchmarks that cover single/multi-image, and short/long video understanding. mPLUG-Owl3 achieves competitive performance with the state-of-the-art methods while reducing inference time and memory usage by 87.8\% and 48.5\% in average. Moreover, we propose a Distractor Resistance evaluation to assess the ability of models to maintain focus amidst distractions. mPLUG-Owl3 also demonstrates outstanding performance in distractor resistance on ultra-long visual sequence inputs. We hope that mPLUG-Owl3 can contribute to the development of more efficient and powerful multimodal large language models.

Endowing large language models (LLMs) with continual learning (CL) capacities is practically important, which enables them to dynamically acquire new knowledge over time. Although many effective methods have been proposed for CL of LLMs, they did not consider online scenarios, thereby sharing a common problem: information leakage (IL), where the task-related information of learned tasks is accessed or reused again. IL not only imposes potential risks on data privacy protection but also significantly hinders the deployment of LLMs in real-world scenarios. To avoid IL while maintaining outstanding CL performance, we propose a novel CL method for LLMs, which first characterizes a parameter-efficient fine-tuning (PEFT) block by a presentative feature distribution, and then dynamically selects the appropriate PEFT blocks for each instance based on its similarity with the presentative feature distributions. Extensive experiments validate the effectiveness of our method on the CL of LLM, showcasing its potential to enhance both privacy and adaptability in practical applications.

Recently, reasoning-based MLLMs have achieved a degree of success in generating long-form textual reasoning chains. However, they still struggle with complex tasks that necessitate dynamic and iterative focusing on and revisiting of visual regions to achieve precise grounding of textual reasoning in visual evidence. We introduce VLM-R³ (Visual Language Model with Region Recognition, Reasoning, and Refinement ), a framework that equips an MLLM with the ability to (i) decide when additional visual evidence is needed, (ii) determine where to ground within the image, and (iii) seamlessly weave the relevant sub-image content back into an interleaved chain-of-thought. The core of our method is \textbf{Region-Conditioned Reinforcement Policy Optimization (R-GRPO)}, a training paradigm that rewards the model for selecting informative regions, formulating appropriate transformations (e.g. crop, zoom), and integrating the resulting visual context into subsequent reasoning steps. To bootstrap this policy, we compile a modest but carefully curated Visuo-Lingual Interleaved Rationale (VLIR) corpus that provides step-level supervision on region selection and textual justification. Extensive experiments on MathVista, ScienceQA, and other benchmarks show that VLM-R$^3$ sets a new state of the art in zero-shot and few-shot settings, with the largest gains appearing on questions demanding subtle spatial reasoning or fine-grained visual cue extraction.

Recent advancements in large language models (LLMs) have significantly enhanced text generation capabilities, yet evaluating their performance in generative writing remains a challenge. Existing benchmarks primarily focus on generic text generation or limited in writing tasks, failing to capture the diverse requirements of high-quality written contents across various domains. To bridge this gap, we present WritingBench, a comprehensive benchmark designed to evaluate LLMs across 6 core writing domains and 100 subdomains. We further propose a query-dependent evaluation framework that empowers LLMs to dynamically generate instance-specific assessment criteria. This framework is complemented by a fine-tuned critic model for criteria-aware scoring, enabling evaluations in style, format and length. The framework's validity is further demonstrated by its data curation capability, which enables a 7B-parameter model to outperform the performance of GPT-4o in writing. We open-source the benchmark, along with evaluation tools and modular framework components, to advance the development of LLMs in writing.

Self-supervised Multi-modal Contrastive Learning (SMCL) remarkably advances modern Vision-Language Pre-training (VLP) models by aligning visual and linguistic modalities. Due to noises in web-harvested text-image pairs, however, scaling up training data volume in SMCL presents considerable obstacles in terms of computational cost and data inefficiency. To improve data efficiency in VLP, we propose Text-aware Image Mixing (TiMix), which integrates mix-based data augmentation techniques into SMCL, yielding significant performance improvements without significantly increasing computational overhead. We provide a theoretical analysis of TiMix from a mutual information (MI) perspective, showing that mixed data samples for cross-modal contrastive learning implicitly serve as a regularizer for the contrastive loss. The experimental results demonstrate that TiMix exhibits a comparable performance on downstream tasks, even with a reduced amount of training data and shorter training time, when benchmarked against existing methods. This work empirically and theoretically demonstrates the potential of data mixing for data-efficient and computationally viable VLP, benefiting broader VLP model adoption in practical scenarios. Our code is available on https://github.com/chaoyajiang/TiMiX/tree/main.

Chi Chen, Yiyang Du, Zheng Fang, Ziyue Wang, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Ziyue Wang, Chi Chen, Yiqi Zhu, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

With an increasing number of music tracks available online, music recommender systems have become popular and ubiquitous. Previous research indicates that people’s preferences, especially in music, dynamically change with various factors, such as surrounding situations and emotional status. However, few existing public recommendation datasets contain such situation or emotion information. Therefore, we constructed SiTunes, a situational music recommendation dataset with rich physiological and psychological signals. We collected the data through a three-stage user study, including: (1) recorded users’ inherent music preference in a lab setting (Stage 1), (2) recorded physiological and environmental situations by smart wristband devices in users’ daily life, and provided psychological and rating feedback for music recommended by traditional recommenders (Stage 2) and (3) by situation-aware recommenders (Stage 3). The experiments were conducted with strict privacy concerns and ethical approval. The dataset contains over 2000 listening logs from 30 users on over 300 music tracks. SiTunes serves as a valuable resource for future studies on situational recommenders and user understanding in recommendation. The dataset is available at https://github.com/JiayuLi-997/SiTunes_dataset/.

Large Language Models (LLMs) exhibit remarkable In-Context Learning (ICL) ability, where the model learns tasks from prompts consisting of input-output examples. However, the pre-training objectives of LLMs often misalign with ICL objectives. They’re mainly pre-trained with methods like masked language modeling and next-sentence prediction. On the other hand, ICL leverages example pairs to guide the model in generating task-aware responses such as text classification and question-answering tasks. The basic pre-training task-related capabilities can sometimes overshadow or conflict with task-specific subtleties required in ICL. To address this, we propose an In-context learning Ability Decoupler (IAD). The model aims to separate the ICL ability from the general ability of LLMs in the meta-training phase, where the ICL-related parameters are separately tuned to adapt for ICL tasks. Concretely, we first identify the parameters that are suitable for ICL by transference-driven gradient importance. We then propose a new max-margin loss to emphasize the separation of the general and ICL abilities. The loss is defined as the difference between the output of ICL and the original LLM, aiming to prevent the overconfidence of the LLM. By meta-training these ICL-related parameters with max-margin loss, we enable the model to learn and adapt to new tasks with limited data effectively. Experimental results show that IAD’s capability yields state-of-the-art performance on benchmark datasets by utilizing only 30% of the model’s parameters. Ablation study and detailed analysis prove the separation of the two abilities.

In vision-language pre-training (VLP), masked image modeling (MIM) has recently been introduced for fine-grained cross-modal alignment. However, in most existing methods, the reconstruction targets for MIM lack high-level semantics, and text is not sufficiently involved in masked modeling. These two drawbacks limit the effect of MIM in facilitating cross-modal semantic alignment. In this work, we propose a semantics-enhanced cross-modal MIM framework (SemMIM) for vision-language representation learning. Specifically, to provide more semantically meaningful supervision for MIM, we propose a local semantics enhancing approach, which harvest high-level semantics from global image features via self-supervised agreement learning and transfer them to local patch encodings by sharing the encoding space. Moreover, to achieve deep involvement of text during the entire MIM process, we propose a text-guided masking strategy and devise an efficient way of injecting textual information in both masked modeling and reconstruction target acquisition. Experimental results validate that our method improves the effectiveness of the MIM task in facilitating cross-modal semantic alignment. Compared to previous VLP models with similar model size and data scale, our SemMIM model achieves state-of-the-art or competitive performance on multiple downstream vision-language tasks.

In video-text retrieval, most existing methods adopt the dual-encoder architecture for fast retrieval, which employs two individual encoders to extract global latent representations for videos and texts. However, they face challenges in capturing fine-grained semantic concepts. In this work, we propose the UNIFY framework, which learns lexicon representations to capture fine-grained semantics and combines the strengths of latent and lexicon representations for video-text retrieval. Specifically, we map videos and texts into a pre-defined lexicon space, where each dimension corresponds to a semantic concept. A two-stage semantics grounding approach is proposed to activate semantically relevant dimensions and suppress irrelevant dimensions. The learned lexicon representations can thus reflect fine-grained semantics of videos and texts. Furthermore, to leverage the complementarity between latent and lexicon representations, we propose a unified learning scheme to facilitate mutual learning via structure sharing and self-distillation. Experimental results show our UNIFY framework largely outperforms previous video-text retrieval methods, with 4.8% and 8.2% Recall@1 improvement on MSR-VTT and DiDeMo respectively.

Multi-modal large language models (MLLMs) have been shown to efficiently integrate natural language with visual information to handle multi-modal tasks. However, MLLMs still face a fundamental limitation of hallucinations, where they tend to generate erroneous or fabricated information. In this paper, we address hallucinations in MLLMs from a novel perspective of representation learning. We first analyzed the representation distribution of textual and visual tokens in MLLM, revealing two important findings: 1) there is a significant gap between textual and visual representations, indicating unsatisfactory cross-modal representation alignment; 2) representations of texts that contain and do not contain hallucinations are entangled, making it challenging to distinguish them. These two observations inspire us with a simple yet effective method to mitigate hallucinations. Specifically, we introduce contrastive learning into MLLMs and use text with hallucination as hard negative examples, naturally bringing representations of non-hallucinative text and visual samples closer while pushing way representations of non-hallucinating and hallucinative text. We evaluate our method quantitatively and qualitatively, showing its effectiveness in reducing hallucination occurrences and improving performance across multiple benchmarks. On the MMhal-Bench benchmark, our method obtains a 34.66% /29.5% improvement over the baseline MiniGPT-4/LLaVA. Our code is available on https://github.com/X-PLUG/mPLUG-HalOwl/tree/main/hacl.

Multi-modal Large Language Models (MLLMs) have demonstrated impressive instruction abilities across various open-ended tasks. However, previous methods primarily fo-cus on enhancing multi-modal capabilities. In this work, we introduce a versatile multi-modal large language model, mPLUG-Owl2, which effectively leverages modality collab-oration to improve performance in both text and multi-modal tasks. mPLUG-Owl2 utilizes a modularized network design, with the language decoder acting as a universal interface for managing different modalities. Specifically, mPLUG-Owl2 incorporates shared functional modules to facilitate modal-ity collaboration and introduces a modality-adaptive module that preserves modality-specific features. Extensive experi-ments reveal that mPLUG-Owl2 is capable of generalizing both text tasks and multi-modal tasks and achieving state-of-the-art performances with a single generic model. Notably, mPLUG-Owl2 is the first MLLM model that demonstrates the modality collaboration phenomenon in both pure-text and multi-modal scenarios, setting a pioneering path in the development of future multi-modal foundation models.

This work proposes a simple training-free prompt-free approach to leverage large language models (LLMs) for the Chinese spelling correction (CSC) task, which is totally different from all previous CSC approaches.The key idea is to use an LLM as a pure language model in a conventional manner.The LLM goes through the input sentence from the beginning, and at each inference step, produces a distribution over its vocabulary for deciding the next token, given a partial sentence.To ensure that the output sentence remains faithful to the input sentence, we design a minimal distortion model that utilizes pronunciation or shape similarities between the original and replaced characters.Furthermore, we propose two useful reward strategies to address practical challenges specific to the CSC task.Experiments on five public datasets demonstrate that our approach significantly improves LLM performance, enabling them to compete with state-of-the-art domain-general CSC models.

Haowei Liu, Xi Zhang, Haiyang Xu, Yaya Shi, Chaoya Jiang, Ming Yan, Ji Zhang, Fei Huang, Chunfeng Yuan, Bing Li, Weiming Hu. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.

Large Language Model (LLM) agents significantly extend the capabilities of standalone LLMs, empowering them to interact with external tools (e.g., APIs, functions) and complete various tasks in a self-directed fashion.The challenge of tool use demands that LLMs not only understand user queries and generate answers accurately but also excel in task planning, tool invocation, and result summarization.While traditional works focus on training a single LLM with all these capabilities, performance limitations become apparent, particularly with smaller models.To overcome these challenges, we propose a novel approach that decomposes the aforementioned capabilities into a planner, caller, and summarizer.Each component is implemented by a single LLM that focuses on a specific capability and collaborates with others to accomplish the task.This modular framework facilitates individual updates and the potential use of smaller LLMs for building each capability.To effectively train this framework, we introduce a two-stage training paradigm.First, we fine-tune a backbone LLM on the entire dataset without discriminating sub-tasks, providing the model with a comprehensive understanding of the task.Second, the fine-tuned LLM is used to instantiate the planner, caller, and summarizer respectively, which are continually fine-tuned on respective sub-tasks.Evaluation across various tool-use benchmarks illustrates that our proposed multi-LLM framework surpasses the traditional single-LLM approach, highlighting its efficacy and advantages in tool learning.

Charts are important for presenting and explaining complex data relationships.Recently, multimodal large language models (MLLMs) have shown remarkable capabilities in chart understanding.However, the sheer size of these models limits their use in resource-constrained environments.In this paper, we present Tiny-Chart, an efficient MLLM for chart understanding with only 3B parameters.TinyChart overcomes two key challenges in efficient chart understanding: (1) reduce the burden of learning numerical computations through Programof-Thoughts (PoT) learning, which trains the model to generate Python programs for numerical calculations, and (2) reduce lengthy vision feature sequences through Vision Token Merging, which gradually merges most similar vision tokens.Extensive experiments demonstrate that our 3B TinyChart achieves SOTA performance on various chart understanding benchmarks including ChartQA, Chart-to-Text, Chart-to-Table, OpenCQA, and ChartX.It outperforms several chart-understanding MLLMs with up to 13B parameters, and close-sourced MLLM GPT-4V on ChartQA, with higher throughput during inference due to a smaller model scale and more efficient vision encoding.

Existing cross-modal fusion methods for temporal language grounding suffer from issues where the generated cross-modal embeddings are affected by noise in their unimodal representations, thus hindering the expression of interactions between language query and the target video segment. Furthermore, the cross-modal representations contain many irrelevant redundancies, which compromises the quality of cross-modal features and thus interferes with accurate moment localization. To address these drawbacks, we propose a novel CrOss-modaL information-constraineD (COLD) model for temporal language grounding, aiming at learning a robust cross-modal embedding representation devoid of irrelevant redundancies and maximizing the interaction between language query and target video moment. Specifically, our model is built upon the principles of the information bottleneck and features two information-constrained modules from different perspectives: 1) the Cross-modal Highlight Information Bottleneck module is designed to maximize the mutual information between language query and target video moment; 2) the Fusion Information Bottleneck module is introduced to constrain the correlations between the cross-modal representations and the localization labels. Comprehensive experimental results on two public benchmark datasets demonstrate the superiority of the proposed model.

Visual grounding focuses on localizing objects referred to by natural language queries. Existing fully and weakly supervised methods rely on a mass of language queries for training. However, collecting natural language queries corresponding to specific objects by annotators is expensive. To reduce the reliance on human-written queries, we propose a novel unsupervised visual grounding framework named VG-Annotator. Different from the existing unsupervised methods that rely on manually designed rules to link objects and language queries. The key idea of VG-Annotator lies in that vision-language pre-trained (VLP) generation models can be language query annotators. Thanks to the powerful multi-modal understanding ability implicitly learned from large-scale pre-training, we consider stimulating models to explicitly generate appropriate descriptions for specific objects in natural language. To this end, we explore a series of multi-modal instructions to indicate which object should be described. We also introduce a supervised fine-tuning process to teach the vision-language models to follow the instructions. Extensive experiments show that the proposed method obtains high-quality language queries. The visual grounding model trained with the generated queries outperforms state-of-the-art unsupervised methods on five widely used datasets.

Weak diagram analysis abilities of LLMs or Multimodal LLMs greatly limit their application scenarios for scientific academic paper writing. In this work, towards a more versatile copilot for academic paper writing, we mainly focus on strengthening the multi-modal diagram analysis ability of Multimodal LLMs. By parsing Latex source files of academic papers, we carefully build a multi-modal diagram understanding dataset M-Paper. By aligning diagrams in the paper with related paragraphs, we construct professional diagram analysis samples for training and evaluation. M-Paper is the first dataset to support joint comprehension of multiple scientific diagrams, including figures and tables in the format of images or Latex codes. Besides, to better align the copilot with the user's intention, we introduce the 'outline' as the control signal, which could be directly given by the user or revised based on auto-generated ones. Comprehensive experiments with a state-of-the-art Multimodal LLM demonstrate that training on our dataset shows stronger scientific diagram understanding performance. The dataset, code, and model are publicly available at https://github.com/X-PLUG/mPLUG-DocOwl/tree/main/PaperOwl.

Large Vision-Language Models (LVLMs) exhibit remarkable capabilities but struggle with ''hallucinations''-inconsistencies between images and their descriptions. Previous hallucination evaluation studies on LVLMs have identified hallucinations in terms of objects, attributes, and relations but overlooked complex hallucinations that create an entire narrative around a fictional entity. In this paper, we introduce a refined taxonomy of hallucinations, featuring a new category: Event Hallucination. We then utilize advanced LLMs to generate and filter fine-grained hallucinatory data consisting of various types of hallucinations, with a particular focus on event hallucinations, laying the groundwork for integrating discriminative and generative evaluation methods within our universal evaluation framework. The proposed benchmark distinctively assesses LVLMs' ability to tackle a broad spectrum of hallucinations, making it a reliable and comprehensive tool for gauging LVLMs' efficacy in handling hallucinations. We will release our code and data.

Recently, temporal action localization (TAL) methods, especially the weakly-supervised and unsupervised ones, have become a hot research topic. Existing unsupervised methods follow an iterative ''clustering and training'' strategy with diverse model designs during training stage, while they often overlook maintaining consistency between these stages, which is crucial: more accurate clustering results can reduce the noises of pseudolabels and thus enhance model training, while more robust training can in turn enrich clustering feature representation. We identify two critical challenges in unsupervised scenarios: 1. What features should the model generate for clustering? 2. Which pseudolabeled instances from clustering should be chosen for model training? After extensive explorations, we proposed a novel yet simple framework called Consistency-Oriented Progressive high actionness Learning to address these issues. For feature generation, our framework adopts a High Actionness snippet Selection (HAS) module to generate more discriminative global video features for clustering from the enhanced actionness features obtained from a designed Inner-Outer Consistency Network (IOCNet). For pseudolabel selection, we introduces a Progressive Learning With Representative Instances (PLRI) strategy to identify the most reliable and informative instances within each cluster for model training. These three modules, HAS, IOCNet, and PLRI, synergistically improve consistency in model training and clustering performance. Extensive experiments on THUMOS'14 and ActivityNet v1.2 datasets under both unsupervised and weakly-supervised settings demonstrate that our framework achieves the state-of-the-art results.

Mobile device operation tasks are increasingly becoming a popular multi-modal AI application scenario. Current Multi-modal Large Language Models (MLLMs), constrained by their training data, lack the capability to function effectively as operation assistants. Instead, MLLM-based agents, which enhance capabilities through tool invocation, are gradually being applied to this scenario. However, the two major navigation challenges in mobile device operation tasks — task progress navigation and focus content navigation — are difficult to effectively solve under the single-agent architecture of existing work. This is due to the overly long token sequences and the interleaved text-image data format, which limit performance. To address these navigation challenges effectively, we propose Mobile-Agent-v2, a multi-agent architecture for mobile device operation assistance. The architecture comprises three agents: planning agent, decision agent, and reflection agent. The planning agent condenses lengthy, interleaved image-text history operations and screens summaries into a pure-text task progress, which is then passed on to the decision agent. This reduction in context length makes it easier for decision agent to navigate the task progress. To retain focus content, we design a memory unit that updates with task progress by decision agent. Additionally, to correct erroneous operations, the reflection agent observes the outcomes of each operation and handles any mistake accordingly. Experimental results indicate that Mobile-Agent-v2 achieves over a 30% improvement in task completion compared to the single-agent architecture of Mobile-Agent. The code is open-sourced at https://github.com/X-PLUG/MobileAgent.

This paper presents MaVEn, an innovative Multi-granularity Visual Encoding framework designed to enhance the capabilities of Multimodal Large Language Models (MLLMs) in multi-image reasoning. Current MLLMs primarily focus on single-image visual understanding, limiting their ability to interpret and integrate information across multiple images. MaVEn addresses this limitation by combining discrete visual symbol sequences, which abstract coarse-grained semantic concepts, with traditional continuous representation sequences that model fine-grained features. This dual approach bridges the semantic gap between visual and textual data, thereby improving the model's ability to process and interpret information from multiple images effectively. Additionally, we design a dynamic reduction mechanism by for long-sequence continuous features to enhance multi-image processing efficiency. Experimental results demonstrate that MaVEn significantly enhances MLLMs' understanding in complex multi-image scenarios, while also improving performance in single-image contexts.

Knowledge Graph Embedding (KGE) is a critical field aiming to transform the elements of knowledge graphs (KGs) into continuous spaces, offering great potential for structured data representation. In contemporary KGE research, the utilization of either hyperbolic or Euclidean space for knowledge graph Embedding is a common practice. However, knowledge graphs encompass diverse geometric data structures, including chains and hierarchies, whose hybrid nature exceeds the capacity of a single embedding space to capture effectively. This paper introduces a novel and highly effective approach called Unified Geometry Knowledge Graph Embedding (UniGE) to address the challenge of representing diverse geometric data in KGs. UniGE stands out as a novel KGE method that seamlessly integrates KGE in both Euclidean and hyperbolic geometric spaces. We introduce an embedding alignment method and fusion strategy, which harnesses optimal transport techniques and the Wasserstein barycenter method. Furthermore, we offer a comprehensive theoretical analysis to substantiate the superiority of our approach, as evident from a more robust error bound. To substantiate the strength of UniGE, we conducted comprehensive experiments on three benchmark datasets. The results consistently demonstrate that UniGE outperforms state-of-the-art methods, aligning with the conclusions drawn from our theoretical analysis.

Referring expression comprehension aims to align natural language queries with visual scenes, which requires establishing fine-grained correspondence between vision and language. This has important applications in multi-modal reasoning systems. Existing methods typically use text-agnostic visual backbones to extract features independently without considering the specific text input. However, we argue that the extracted visual features can be inconsistent with the referring expression, which hurts multi-modal understanding. To address this, we first propose Query-modulated Refinement Network (QRNet) that leverages language guidance to guide visual feature extraction. However, it only focuses on the grounding task that can only provide coarse-grained annotations in the form of bounding box coordinates. The guidance for the visual backbone is indirect, and the inconsistent issue still exists. To this end, we further propose UniQRNet, a multi-task framework over the QRNet to learn referring expression grounding and segmentation jointly. The framework introduces a multi-task head that leverages fine-grained pixel-level supervision from the segmentation task to directly guide the intermediate layers of QRNet to learn text-consistent visual features. Besides, UniQRNet also includes a loss balance strategy that allows two types of supervision signals to cooperate and optimize the model together. We conduct the most comprehensive comparison experiment covering four major datasets, ten evaluation set and three evaluation metrics used in previous work. UniQRNet outperforms previous state-of-the-art methods by a large margin on both referring comprehensive grounding (1.8%~5.09%) and segmentation tasks (0.57%~5.56%). Ablation and analysis reveal that UniQRNet can improve the consistency of visual features with text input and can bring significant performance improvement.

In open-domain dialogue generation tasks, contexts and responses in most datasets are oneto-one mapped, violating an important manyto-many characteristic: a context leads to various responses, and a response answers multiple contexts.Without such patterns, models poorly generalize and prefer responding safely.Many attempts have been made in either multiturn settings from a one-to-many perspective or in a many-to-many perspective but limited to single-turn settings.The major challenge to many-to-many augment multi-turn dialogues is that discretely replacing each turn with semantic similarity breaks fragile context coherence.In this paper, we propose DialoGue Path Sampling (DialoGPS) method in continuous semantic space, the first many-to-many augmentation method for multi-turn dialogues.Specifically, we map a dialogue to our extended Brownian Bridge, a special Gaussian process.We sample latent variables to form coherent dialogue paths in the continuous space.A dialogue path corresponds to a new multi-turn dialogue and is used as augmented training data.We show the effect of DialoGPS with both automatic and human evaluation.

Video-language pre-training has advanced the performance of various downstream video-language tasks. However, most previous methods directly inherit or adapt typical image-language pre-training paradigms to video-language pre-training, thus not fully exploiting the unique characteristic of video, i.e., temporal. In this paper, we propose a Hierarchical Temporal-Aware video-language pre-training framework, HiTeA, with two novel pre-training tasks for yielding temporal-aware multi-modal representation with cross-modal fine-grained temporal moment information and temporal contextual relations between video-text multi-modal pairs. First, we propose a cross-modal moment exploration task to explore moments in videos by mining the paired texts, which results in detailed video moment representation. Then, based on the learned detailed moment representations, the inherent temporal contextual relations are captured by aligning video-text pairs as a whole in different time resolutions with multi-modal temporal relation exploration task. Furthermore, we introduce the shuffling test to evaluate the temporal reliance of datasets and video-language pre-training models. We achieve state-of-the-art results on 15 well-established video-language understanding and generation tasks, especially on temporal-oriented datasets (e.g., SSv2-Template and SSv2-Label) with 8.6% and 11.1% improvement respectively. HiTeA also demonstrates strong generalization ability when directly transferred to downstream tasks in a zero-shot manner.

Recent years have witnessed a big convergence of language, vision, and multi-modal pretraining. In this work, we present mPLUG-2, a new unified paradigm with modularized design for multi-modal pretraining, which can benefit from modality collaboration while addressing the problem of modality entanglement. In contrast to predominant paradigms of solely relying on sequence-to-sequence generation or encoder-based instance discrimination, mPLUG-2 introduces a multi-module composition network by sharing common universal modules for modality collaboration and disentangling different modality modules to deal with modality entanglement. It is flexible to select different modules for different understanding and generation tasks across all modalities including text, image, and video. Empirical study shows that mPLUG-2 achieves state-of-the-art or competitive results on a broad range of over 30 downstream tasks, spanning multi-modal tasks of image-text and video-text understanding and generation, and uni-modal tasks of text-only, image-only, and video-only understanding. Notably, mPLUG-2 shows new state-of-the-art results of 48.0 top-1 accuracy and 80.3 CIDEr on the challenging MSRVTT video QA and video caption tasks with a far smaller model size and data scale. It also demonstrates strong zero-shot transferability on vision-language and video-language tasks. Code and models will be released in https://github.com/X-PLUG/mPLUG-2.

Vision-Language Pre-training (VLP) methods based on object detection enjoy the rich knowledge of fine-grained object-text alignment but at the cost of computationally expensive inference. Recent Visual-Transformer (ViT)-based approaches circumvent this issue while struggling with long visual sequences without detailed cross-modal alignment information. This paper introduces a ViT-based VLP technique that efficiently incorporates object information through a novel patch-text alignment mechanism. Specifically, we convert object-level signals into patch-level ones and devise a Patch-Text Alignment pre-training task (PTA) to learn a text-aware patch detector. By using off-the-shelf delicate object annotations in 5% training images, we jointly train PTA with other conventional VLP objectives in an end-to-end manner, bypassing the high computational cost of object detection and yielding an effective patch detector that accurately detects text-relevant patches, thus considerably reducing patch sequences and accelerating computation within the ViT backbone. Our experiments on a variety of widely-used benchmarks reveal that our method achieves a speedup of nearly 88% compared to prior VLP models while maintaining competitive or superior performance on downstream tasks with similar model size and data scale.

Previous dual-encoder pre-training methods for video-text retrieval employ contrastive learning for cross-modal alignment in a latent space. However, such learned latent spaces often result in modality gap problem [26]. In this paper, we introduce a novel SemVTR framework designed to learn semantics-grounded video-text representations in a vocabulary space, in which each dimension corresponds to a semantic concept represented by a word. The representation is obtained by grounding video and text into semantically-related dimensions with high activation values. As video-text pairs share grounded dimensions, their vocabulary representations are expected to cluster together and thus alleviate modality gap problem. So, the crux of our method lies in grounding video and text into vocabulary space. Specifically, we propose a Multi-Granularity Video Semantics Grounding approach and a Textual Semantics Preserving training strategy. The visualization illustrates that SemVTR obtains semantics-gronded vocabulary representation and also alleviates the modality gap problem. SemVTR significantly outperforms existing methods on four video-text retrieval benchmarks.

Inspired by the recent developments of large language models (LLMs), we propose mPLUG-Octopus, a versatile conversational assistant designed to provide users with coherent, engaging, and helpful interaction experiences in both text-only and multi-modal scenarios. Unlike traditional pipeline chatting systems, mPLUG-Octopus offers a diverse range of creative capabilities including open-domain QA, multi-turn chatting, and multi-modal creation, all built with a unified multimodal LLM without relying on any external API. With the modularized end-to-end multimodal LLM technology, mPLUG-Octopus efficiently facilitates engaging and open-domain conversation experience. It exhibits a wide range of uni/multi-modal elemental capabilities, enabling it to seamlessly communicate with users on open-domain topics and engage in multi-turn conversations. It also assists users in accomplishing various content creation and application tasks. Our conversational assistant can also be deployed on smart hardware to drive advanced AIGC applications.

The Visual Question Answering (VQA) task utilizes both visual image and language analysis to answer a textual question with respect to an image. It has been a popular research topic with an increasing number of real-world applications in the last decade. This paper introduces a novel hierarchical integration of vision and language AliceMind-MMU (ALIbaba’s Collection of Encoder-decoders from Machine IntelligeNce lab of Damo academy - MultiMedia Understanding) , which leads to similar or even slightly better results than a human being does on VQA. A hierarchical framework is designed to tackle the practical problems of VQA in a cascade manner including: (1) diverse visual semantics learning for comprehensive image content understanding; (2) enhanced multi-modal pre-training with modality adaptive attention; and (3) a knowledge-guided model integration with three specialized expert modules for the complex VQA task. Treating different types of visual questions with corresponding expertise needed plays an important role in boosting the performance of our VQA architecture up to the human level. An extensive set of experiments and analysis are conducted to demonstrate the effectiveness of the new research work.

Customer reviews usually contain much information about one's online shopping experience. While positive reviews are beneficial to the stores, negative ones will largely influence consumers' decision and may lead to a decline in sales. Therefore, it is of vital importance to carefully and persuasively reply to each negative review and minimize its disadvantageous effect. Recent studies consider leveraging generation models to help the sellers respond. However, this problem is not well-addressed as the reviews may contain multiple aspects of issues which should be resolved accordingly and persuasively. In this work, we propose a Multi-Source Multi-Aspect Attentive Generation model for persuasive response generation. Various sources of information are appropriately obtained and leveraged by the proposed model for generating more informative and persuasive responses. A multi-aspect attentive network is proposed to automatically attend to different aspects in a review and ensure most of the issues are tackled. Extensive experiments on two real-world datasets, demonstrate that our approach outperforms the state-of-the-art methods and online tests prove that our deployed system significantly enhances the efficiency of the stores' dealing with negative reviews.

Intent detection is at the core of task-oriented dialogue systems. Existing intent detection systems are typically trained with a large amount of data over a predefined set of intent classes. However, newly emerged intents in multiple domains are commonplace in the real world. And it is time-consuming and impractical for dialogue systems to re-collect enough annotated data and re-train the model. These limitations call for an intent detection system that could continually recognize new intents with very few labeled examples. In this work, we study the Continual Few-shot Intent Detection (CFID) problem and construct a benchmark consisting of nine tasks with multiple domains and imbalanced classes. To address the key challenges of (a) catastrophic forgetting during continuous learning and (b) negative knowledge transfer across tasks, we propose the Prefix-guided Lightweight Encoder (PLE) with three auxiliary strategies, namely Pseudo Samples Replay (PSR), Teacher Knowledge Transfer (TKT) and Dynamic Weighting Replay (DWR). Extensive experiments demonstrate the effectiveness and efficiency of our method in preventing catastrophic forgetting and encouraging positive knowledge transfer across tasks.

Visual grounding focuses on establishing fine-grained alignment between vision and natural language, which has essential applications in multimodal reasoning systems. Existing methods use pre-trained query-agnostic visual backbones to extract visual feature maps independently without considering the query information. We argue that the visual features extracted from the visual backbones and the features really needed for multimodal reasoning are inconsistent. One reason is that there are differences between pre-training tasks and visual grounding. Moreover, since the backbones are query-agnostic, it is difficult to completely avoid the inconsistency issue by training the visual backbone end-to-end in the visual grounding framework. In this paper, we propose a Query-modulated Refinement Network (QRNet) to address the inconsistent issue by adjusting intermediate features in the visual backbone with a novel Query-aware Dynamic Attention (QD-ATT) mechanism and query-aware multiscale fusion. The QD-ATT can dynamically compute query-dependent visual attention at the spatial and channel levels of the feature maps produced by the visual backbone. We apply the QRNet to an end-to-end visual grounding framework. Extensive experiments show that the proposed method outperforms state-of-the-art methods on five widely used datasets. Our code is available at https://github.com/LukeForeverYoung/QRNet.

Chenliang Li, Haiyang Xu, Junfeng Tian, Wei Wang, Ming Yan, Bin Bi, Jiabo Ye, He Chen, Guohai Xu, Zheng Cao, Ji Zhang, Songfang Huang, Fei Huang, Jingren Zhou, Luo Si. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022.

Multimodal named entity recognition (MNER) aims to detect and classify named entities in multimodal scenarios. It requires bridging the gap between natural language and visual context, which presents two-fold challenges: the cross-modal alignment is diversified, and the cross-modal interaction is sometimes implicit. Existing MNER methods are vulnerable to some implicit interactions and are prone to overlook the involved significant features. To tackle this problem, we novelly propose to refine the cross-modal attention by identifying and highlighting some task-salient features. The saliency of each feature is measured according to its correlation with the expanded entity label words derived from external knowledge bases. We further propose an end-to-end Transformer-based MNER framework, which holds neater architecture yet achieves better performance than previous methods. Extensive experiments are conducted to validate the merits of our method. Moreover, our method reveals a significant advantage in data efficiency and generalization ability.

Although pre-trained language models (PLMs) have achieved state-of-the-art performance on various natural language processing (NLP) tasks, they are shown to be lacking in knowledge when dealing with knowledge driven tasks. Despite the many efforts made for injecting knowledge into PLMs, this problem remains open. To address the challenge, we propose DictBERT, a novel approach that enhances PLMs with dictionary knowledge which is easier to acquire than knowledge graph (KG). During pre-training, we present two novel pre-training tasks to inject dictionary knowledge into PLMs via contrastive learning: dictionary entry prediction and entry description discrimination. In fine-tuning, we use the pre-trained DictBERT as a plugin knowledge base (KB) to retrieve implicit knowledge for identified entries in an input sequence, and infuse the retrieved knowledge into the input to enhance its representation via a novel extra-hop attention mechanism. We evaluate our approach on a variety of knowledge driven and language understanding tasks, including NER, relation extraction, CommonsenseQA, OpenBookQA and GLUE. Experimental results demonstrate that our model can significantly improve typical PLMs: it gains a substantial improvement of 0.5%, 2.9%, 9.0%, 7.1% and 3.3% on BERT-large respectively, and is also effective on RoBERTa-large.

Video domain adaptation aims to transfer knowledge from labeled source videos to unlabeled target videos. Existing video domain adaptation methods require full access to the source videos to reduce the domain gap between the source and target videos, which are impractical in real scenarios where the source videos are not available with concerns in transmission efficiency or privacy issues. To address this problem, in this paper, we propose to solve a source-free domain adaptation task for videos where only a pre-trained source model and unlabeled target videos are available for learning a multimodal video classification model. Existing source-free domain adaptation methods cannot be directly applied to this task, since videos always suffer from domain discrepancy along both the multimodal and temporal aspects, which brings difficulties in domain adaptation especially when the source data are unavailable. In this paper, we propose a Multimodal and Temporal Relative Alignment Network (MTRAN) to deal with the above challenges. To explicitly imitate the domain shifts contained in the multimodal information and the temporal dynamics of the source and target videos, we divide the target videos into two splits according to the self-entropy values of the classification results. The low-entropy videos are deemed to be source-like while the high-entropy videos are deemed to be target-like. Then, we adopt a self-entropy-guided MixUp strategy to generate synthetic samples and hypothetical samples as instance-level based on source-like and target-like videos, and push each synthetic sample to be similar with the corresponding hypothetical sample that is slightly closer to the source-like videos than the synthetic sample by multimodal and temporal relative alignment schemes. We evaluate the proposed model on four public video datasets. The results show that our model outperforms existing state-of-the-art methods.

Video-text retrieval has been a crucial and fundamental task in multi-modal research. The development of video-text retrieval has been considerably promoted by large-scale multi-modal contrastive pre-training, which primarily focuses on coarse-grained or fine-grained contrast. However, cross-grained contrast, which is the contrast between coarse-grained representations and fine-grained representations, has rarely been explored in prior research. Compared with fine-grained or coarse-grained contrasts, cross-grained contrast calculate the correlation between coarse-grained features and each fine-grained feature, and is able to filter out the unnecessary fine-grained features guided by the coarse-grained feature during similarity calculation, thus improving the accuracy of retrieval. To this end, this paper presents a novel multi-grained contrastive model, namely X-CLIP, for video-text retrieval. However, another challenge lies in the similarity aggregation problem, which aims to aggregate fine-grained and cross-grained similarity matrices to instance-level similarity. To address this challenge, we propose the Attention Over Similarity Matrix (AOSM) module to make the model focus on the contrast between essential frames and words, thus lowering the impact of unnecessary frames and words on retrieval results. With multi-grained contrast and the proposed AOSM module, X-CLIP achieves outstanding performance on five widely-used video-text retrieval datasets, including MSR-VTT (49.3 [email protected]), MSVD (50.4 [email protected]), LSMDC (26.1 [email protected]), DiDeMo (47.8 [email protected]) and ActivityNet (46.2 [email protected]).

Composed Query Based Image Retrieval (CQBIR) aims at searching images relevant to a composed query, i.e., a reference image together with a modifier text. Compared with conventional image retrieval, which takes a single image or text to retrieve desired images, CQBIR encounters more challenges as it requires not only effective semantic correspondence between the heterogeneous query and target, but also synergistic understanding of the composed query. To establish robust CQBIR model, four critical types of relational information can be included, i.e., cross-modal, intra-sample, inter-sample, and cross-sample relationships. Pioneer studies mainly exploit parts of the information, which are hard to make them enhance and complement each other. In this paper, we propose a comprehensive relationship reasoning network by fully exploring the four types of information for CQBIR, which mainly includes two key designs. First, we introduce a memory-augmented cross-modal attention module, in which the representation of the composed query is augmented by considering the cross-modal relationship between the reference image and the modification text. Second, we design a multi-scale matching strategy to optimize our network, aiming at harnessing information from the intra-sample, inter-sample, and cross-sample relationships. To the best of our knowledge, this is the first work to fully explore the four pieces of relationships in a unified deep model for CQBIR. Comprehensive experimental results on five standard benchmarks demonstrate that the proposed method performs favorably against state-of-the-art models.

Jianhai Zhang, Mieradilijiang Maimaiti, Gao Xing, Yuanhang Zheng, Ji Zhang. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2022.

In multilingual communities, code-switching is a common phenomenon and code-switched tasks have become a crucial area of research in natural language processing (NLP) applications. Existing approaches mainly focus on supervised learning. However, it is expensive to annotate a sufficient amount of code-switched data. In this paper, we consider zero-shot setting and improve model performance on code-switched tasks via monolingual language datasets, unlabeled code-switched datasets, and semantic dictionaries. Inspired by the mechanism of code-switching itself, we propose multi-label masked language modeling and predict both the masked word and its synonyms in other languages. Experimental results show that compared with baselines, our method can further improve the pretrained multilingual model's performance on code-switched sentiment analysis datasets.

Qianglong Chen, Feng Ji, Xiangji Zeng, Feng-Lin Li, Ji Zhang, Haiqing Chen, Yin Zhang. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

Existing data-driven methods can well handle short text generation. However, when applied to the long-text generation scenarios such as story generation or advertising text generation in the commercial scenario, these methods may generate illogical and uncontrollable texts. To address these aforementioned issues, we propose a graph-based grouping planner~(GGP) following the idea of first-plan-then-generate. Specifically, given a collection of key phrases, GGP firstly encodes these phrases into a instance-level sequential representation and a corpus-level graph-based representation separately. With these two synergic representations, we then regroup these phrases into a fine-grained plan, based on which we generate the final long text. We conduct our experiments on three long text generation datasets and the experimental results reveal that GGP significantly outperforms baselines, which proves that GGP can control the long text generation with knowing how to say and in what order.

Knowledge enhanced pre-trained language models (K-PLMs) are shown to be effective for many public tasks in the literature, but few of them have been successfully applied in practice. To address this problem, we propose K-AID, a systematic approach that includes a low-cost knowledge acquisition process for acquiring domain knowledge, an effective knowledge infusion module for improving model performance, and a knowledge distillation component for reducing the model size and deploying K-PLMs on resource-restricted devices (e.g., CPU) for real-world application. Importantly, instead of capturing entity knowledge like the majority of existing K-PLMs, our approach captures relational knowledge, which contributes to better improving sentence-level text classification and text matching tasks that play a key role in question answering (QA). We conducted a set of experiments on five text classification tasks and three text matching tasks from three domains, namely E-commerce, Government, and Film&TV, and performed online A/B tests in E-commerce. Experimental results show that our approach is able to achieve substantial improvement on sentence-level question answering tasks and bring beneficial business value in industrial settings.

Live streaming is becoming an increasingly popular trend of sales in E-commerce. The core of live-streaming sales is to encourage customers to purchase in an online broadcasting room. To enable customers to better understand a product without jumping out, we propose AliMe MKG, a multi-modal knowledge graph that aims at providing a cognitive profile for products, through which customers are able to seek information about and understand a product. Based on the MKG, we build an online live assistant that highlights product search, product exhibition and question answering, allowing customers to skim over item list, view item details, and ask item-related questions. Our system has been launched online in the Taobao app, and currently serves hundreds of thousands of customers per day.

Mieradilijiang Maimaiti, Yang Liu, Yuanhang Zheng, Gang Chen, Kaiyu Huang, Ji Zhang, Huanbo Luan, Maosong Sun. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021.

A number of studies point out that current Visual Question Answering (VQA) models are severely affected by the language prior problem, which refers to blindly making predictions based on the language shortcut. Some efforts have been devoted to overcoming this issue with delicate models. However, there is no research to address it from the view of the answer feature space learning, despite the fact that existing VQA methods all cast VQA as a classification task. Inspired by this, in this work, we attempt to tackle the language prior problem from the viewpoint of the feature space learning. An adapted margin cosine loss is designed to discriminate the frequent and the sparse answer feature space under each question type properly. In this way, the limited patterns within the language modality can be largely reduced to eliminate the language priors. We apply this loss function to several baseline models and evaluate its effectiveness on two VQA-CP benchmarks. Experimental results demonstrate that our proposed adapted margin cosine loss can enhance the baseline models with an absolute performance gain of 15\% on average, strongly verifying the potential of tackling the language prior problem in VQA from the angle of the answer feature space learning.

Vision-and-language pretraining (VLP) aims to learn generic multimodal representations from massive image-text pairs. While various successful attempts have been proposed, learning fine-grained semantic alignments between image-text pairs plays a key role in their approaches. Nevertheless, most existing VLP approaches have not fully utilized the intrinsic knowledge within the image-text pairs, which limits the effectiveness of the learned alignments and further restricts the performance of their models. To this end, we introduce a new VLP method called ROSITA, which integrates the cross- and intra-modal knowledge in a unified scene graph to enhance the semantic alignments. Specifically, we introduce a novel structural knowledge masking (SKM) strategy to use the scene graph structure as a priori to perform masked language (region) modeling, which enhances the semantic alignments by eliminating the interference information within and across modalities. Extensive ablation studies and comprehensive analysis verifies the effectiveness of ROSITA in semantic alignments. Pretrained with both in-domain and out-of-domain datasets, ROSITA significantly outperforms existing state-of-the-art VLP methods on three typical vision-and-language tasks over six benchmark datasets.

We present AliMe Avatar, a Vtuber designed for live-streaming sales in the E-commerce field. To support the emerging live shopping mode, the core of our digitial avatar is to enable customers to understand products and encourage customers to purchase in a virtual broadcasting room. Based on computer graphics & vision, natural language processing, and speech recognition & synthesis, our AI avatar is able to offer three kinds of key capabilities: custom appearance, product broadcasting, and multi-modal interaction. Currently, it has been launched online in the Taobao app, broadcasts 700+ hours and serves hundreds of thousands of customers per day. In this paper, we mainly focus on the product broadcasting part, demonstrate the system, present the underlying techniques, and share our experience in dealing with live-streaming E-commerce.

Cold-start is the most difficult and time-consuming phase when building a question answering based chatbot for a new business scenario because of the collection of sufficient training data. In this paper, we propose AliMe DA, a practical data augmentation (DA) framework that consists of data production, denoising and consumption, to alleviate this problem. We show how our DA approach can be used to substantially enhance annotation productivity and also improve downstream model performance. More importantly, we provide best practices for data augmentation, including how to choose and employ appropriate methods at each stage of our framework, and share our observation on the applicable scene of data augmentation in the era of pre-trained language models.

Given a user query, traditional multi-turn retrieval-based dialogue systems first retrieve a set of candidate responses from the historical dialogue sessions. Then the response selection models select the most appropriate response to the given query. However, previous work only considers the matching between the query and the response but ignores the informative dialogue session in which the response is located. Nevertheless, this session, composed of the response, the response's history and the response's future, always contains valuable contextual information which can help the response selection task. More specifically, if the current query and a response's history both refer to the same question, we can conclude that this response is quite likely to answer this query. As for the response's future, it can always provide contextual hints and supplementary information that might be omitted in the response. Inspired by such motivation, we propose a query-to-session matching (QSM) framework to make full use of the session information: matching the query with the candidate session instead of the response only. Different from the previous work which ranks response directly, the response in the session with the highest query-to-session matching score will be selected as the desired response. In our proposed framework, the query, history, and future are all sequences of utterances, which makes it necessary to model the relationships among the utterances. So we propose a novel dialogue flow aware query-to-session matching (DF-QSM) model. The dialogue flows model the relationships among the utterances through a memory network. To our best knowledge, our paper is the first work to utilize both the response's history and future in the response selection task. The experimental results on three multi-turn response selection benchmarks show that our proposed model outperforms existing state-of-the-art methods by a large margin.

Pre-­sales customer service is of importance to E­-commerce plat­forms as it contributes to optimizing customers? buying process. To better serve users, we propose AliMe KG, a domain knowledge graph in E­-commerce that captures user problems, points of inter­est (POI), item information and relations thereof. It helps to under­ stand user needs, answer pre­-sales questions and generate explana­tion texts. We applied AliMe KG to several online business scenar­ios such as shopping guide, question answering over properties and selling point generation, and gained positive and beneficial business results. In the paper, we systematically introduce how we construct domain knowledge graph from free text, and demonstrate its busi­ness value with several applications. Our experience shows that min­ ing structured knowledge from free text in vertical domain is prac­ticable, and can be of substantial value in industrial settings.

Visual storytelling aims to generate a narrative paragraph from a sequence of images automatically.Existing approaches construct text description independently for each image and roughly concatenate them as a story, which leads to the problem of generating semantically incoherent content.In this paper, we propose a new way for visual storytelling by introducing a topic description task to detect the global semantic context of an image stream.A story is then constructed with the guidance of the topic description.In order to combine the two generation tasks, we propose a multi-agent communication framework that regards the topic description generator and the story generator as two agents and learn them simultaneously via iterative updating mechanism.We validate our approach on VIST dataset, where quantitative results, ablations, and human evaluation demonstrate our method's good ability in generating stories with higher quality compared to state-of-the-art methods.

A fundamental trade-off between effectiveness and efficiency needs to be balanced when designing an online question answering system. Effectiveness comes from sophisticated functions such as extractive machine reading comprehension (MRC), while efficiency is obtained from improvements in preliminary retrieval components such as candidate document selection and paragraph ranking. Given the complexity of the real-world multi-document MRC scenario, it is difficult to jointly optimize both in an end-to-end system. To address this problem, we develop a novel deep cascade learning model, which progressively evolves from the documentlevel and paragraph-level ranking of candidate texts to more precise answer extraction with machine reading comprehension. Specifically, irrelevant documents and paragraphs are first filtered out with simple functions for efficiency consideration. Then we jointly train three modules on the remaining texts for better tracking the answer: the document extraction, the paragraph extraction and the answer extraction. Experiment results show that the proposed method outperforms the previous state-of-the-art methods on two large-scale multidocument benchmark datasets, i.e., TriviaQA and DuReader. In addition, our online system can stably serve typical scenarios with millions of daily requests in less than 50ms.

Existing approaches to neural machine translation are typically autoregressive models.While these models attain state-of-the-art translation quality, they are suffering from low parallelizability and thus slow at decoding long sequences.In this paper, we propose a novel model for fast sequence generationthe semi-autoregressive Transformer (SAT).The SAT keeps the autoregressive property in global but relieves in local and thus are able to produce multiple successive words in parallel at each time step.Experiments conducted on English-German and Chinese-English translation tasks show that the SAT achieves a good balance between translation quality and decoding speed.On WMT'14 English-German translation, the SAT achieves 5.58× speedup while maintaining 88% translation quality, significantly better than the previous non-autoregressive methods.When produces two words at each time step, the SAT is almost lossless (only 1% degeneration in BLEU score).