NJNing Jiang

Ning Jiang

*This is an algorithmically generated author record based on publicly available information.

Ning Jiang has a total of 89 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    FailSafe-RAG: Evidence-State-Aware Generation under Retrieval Failures

    by Wanjie Tao, Qun Dai, Quan Lu, Ning Jiang, Zulong Chen

  • Open Access

    Group Behavior-Guided Multi-Density Graph Clustering for Large-Scale Telemarketing Recommendation

    by Ke Wang, Wanjie Tao, Yuchen Yuan, Zhang-Hua Fu, Xin Cai, Yanyan Zheng, Quan Lu, Ning Jiang, Yuanyuan Liu

    Large-scale telemarketing recommendation is crucial for personalized customer outreach in various industries like banking and telecommunications. However, traditional individual-level recommendation approaches suffer from high computational overhead in large-scale scenarios, making group-level recommendation a promising alternative. Nevertheless, existing group-level recommendation methods rely on distance-based similarity metrics that overlook long-tail density variations and decouple clustering from resource allocation, leading to suboptimal allocation decisions. To address these challenges, we propose a novel Group Behavior-guided Multi-density Graph Clustering (GBMGC) framework, which integrates multi-density behavior-aware graph clustering with constrained resource allocation optimization for efficient and scalable telemarketing recommendation. The framework adopts a two-stage design consisting of Density-aware Graph Clustering (DGC) and Group-guided Optimization Recommendation (GOR), with Tree-structured Parzen Estimator (TPE)-based hyperparameter selection. Specifically, to overcome the limitation of conventional distance-based similarity grouping in capturing long-tail behavioral density patterns, DGC constructs a weighted behavioral graph using heuristic weights derived from statistical tests on historical response distributions and applies adaptive density thresholds to identify groups with different population densities. In the second stage, based on the resulting group structure, GOR formulates the scheduling process as a constrained integer programming problem to maximize expected answer rates under practical constraints. Finally, TPE selects the key hyperparameters of DGC and GOR using validation-set performance as feedback, reducing manual hyperparameter search. Evaluated on a financial telemarketing test dataset with more than 3 M records and the MovieLens-1 M test dataset with approximately 0.2 M records, GBMGC achieves higher answer rate (AR) point estimates than the compared methods on the historically matched observations under the adopted offline evaluation protocol, while maintaining favorable computational efficiency. The results on MovieLens-1 M provide additional evidence of its applicability beyond the telemarketing recommendation setting.

  • Open Access

    NBA-Net: Next-Behavior-Aware Network for Intent Recommendation

    by Haoxin Shen, Peng Ying, Wanjie Tao, Jie Liang, Quan Lu, Ning Jiang

  • Open Access

    Rethinking MoE with Retrieval-Memory Synergy: Towards Efficient Expert Coordination

    by Wanjie Tao, Qun Dai, Yantong Lv, Quan Lu, Ning Jiang, Zulong Chen

  • Open Access

    Referring Expression Instance Retrieval and A Strong End-to-End Baseline

    by Xiangzhao Hao, Kuan Zhu, Hongyu Guo, Haiyun Guo, Ning Jiang, Quan Lu, Ming Tang, Jinqiao Wang


  • PMMTalk$:$ Speech-Driven 3D Facial Animation From Complementary Pseudo Multi-Modal Features

    by Tianshun Han, Shengnan Gui, Yiqing Huang, Baihui Li, Lijian Liu, Benjia Zhou, Ning Jiang, Quan Lu, Ruicong Zhi, Yanyan Liang, et al.

    Speech-driven 3D facial animation has improved a lot recently while most related works only utilize acoustic modality and neglect the influence of visual and textual cues, leading to unsatisfactory results in terms of precision and coherence. We argue that visual and textual cues are not trivial information. Therefore, we present a novel framework, namely PMMTalk, using complementary P seudo M ulti- M odal features for improving the accuracy of facial animation. The framework entails three modules: PMMTalk encoder, cross-modal alignment module, and PMMTalk decoder. Specifically, the PMMTalk encoder employs the off-the-shelf talking head generation architecture and speech recognition technology to extract visual and textual information from speech, respectively. Following this, the cross-modal alignment module aligns the audio-image-text features at temporal and semantic levels. Subsequently, the PMMTalk decoder is employed to predict lip-syncing facial blendshape coefficients. Contrary to prior methods, PMMTalk only requires an additional random reference face image but yields more accurate results. Additionally, it is artist-friendly as it seamlessly integrates into standard animation production workflows by introducing facial blendshape coefficients. Finally, given the scarcity of 3D talking face datasets, we introduce a large-scale 3D C hinese A udio- V isual F acial A nimation (3D-CAVFA) dataset. Extensive experiments and user studies show that our approach outperforms the state of the art. Codes and datasets are available at PMMTalk.


  • Open Access

    Learning Individual Treatment Effects under Heterogeneous Interference in Networks

    by Ziyu Zhao, Yuqi Bai, Ruoxuan Xiong, Qingyu Cao, Chao Ma, Ning Jiang, Fei Wu, Kun Kuang

    Estimating individual treatment effects in networked observational data is a crucial and increasingly recognized problem. One major challenge of this problem is violating the stable unit treatment value assumption (SUTVA), which posits that a unit’s outcome is independent of others’ treatment assignments. However, in network data, a unit’s outcome is influenced not only by its treatment (i.e., direct effect) but also by the treatments of others (i.e., spillover effect) since the presence of interference. Moreover, the interference from other units is always heterogeneous (e.g., friends with similar interests have a different influence than those with different interests). In this article, we focus on the problem of estimating individual treatment effects (including direct effect and spillover effect) under heterogeneous interference in networks. To address this problem, we propose a novel dual weighting regression (DWR) algorithm by simultaneously learning attention weights to capture the heterogeneous interference from neighbors and sample weights to eliminate the complex confounding bias in networks. We formulate the learning process as a bi-level optimization problem. Theoretically, we give a generalization error bound for the expected estimation error of the individual treatment effects. Extensive experiments on four benchmark datasets demonstrate that the proposed DWR algorithm outperforms the state-of-the-art methods in estimating individual treatment effects under heterogeneous network interference.


  • Open Access

    Conversational Speech Recognition by Learning Audio-textual Cross-modal Contextual Representation

    by Kun Wei, Bei Li, Hang Lv, Quan Lu, Ning Jiang, Lei Xie

    Automatic Speech Recognition (ASR) in conversational settings presents unique challenges, including extracting relevant contextual information from previous conversational turns. Due to irrelevant content, error propagation, and redundancy, existing methods struggle to extract longer and more effective contexts. To address this issue, we introduce a novel conversational ASR system, extending the Conformer encoder-decoder model with cross-modal conversational representation. Our approach leverages a cross-modal extractor that combines pre-trained speech and text models through a specialized encoder and a modal-level mask input. This enables the extraction of richer historical speech context without explicit error propagation. We also incorporate conditional latent variational modules to learn conversational level attributes such as role preference and topic coherence. By introducing both cross-modal and conversational representations into the decoder, our model retains context over longer sentences without information loss, achieving relative accuracy improvements of 8.8% and 23% on Mandarin conversation datasets HKUST and MagicData-RAMC, respectively, compared to the standard Conformer model.

  • Conversational Speech Recognition by Learning Audio-Textual Cross-Modal Contextual Representation

    by Kun Wei, Bei Li, Hang Lv, Quan Lu, Ning Jiang, Lei Xie

    Automatic Speech Recognition (ASR) in conversational settings presents unique challenges, including extracting relevant contextual information from previous conversational turns. Due to irrelevant content, error propagation, and redundancy, existing methods struggle to extract longer and more effective contexts. To address this issue, we introduce a novel conversational ASR system, extending the Conformer encoder-decoder model with cross-modal conversational representation. Our approach leverages a cross-modal extractor that combines pre-trained speech and text models through a specialized encoder and a modal-level mask input. This enables the extraction of richer historical speech context without explicit error propagation. We also incorporate conditional latent variational modules to learn conversational-level attributes such as role preference and topic coherence. By introducing both cross-modal and conversational representations into the decoder, our model retains longer context without information loss, achieving relative accuracy improvements of 8.8% and 23% on Mandarin conversation datasets HKUST and MagicData-RAMC, respectively, compared to the standard Conformer model.