Junqi Liao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Junqi Liao has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Xin Wei
Nanjing University of Posts and Telecommunications
6 shared publications - Liang Zhou
Key Laboratory of Broadband Wireless Communication and Sensor Network Technology (Ministry of Education), Nanjing University of Posts and Telecommunications, Nanjing, China
4 shared publications Bangbang Hou
2 shared publicationsDan Wu
2 shared publicationsJie Song
2 shared publications- Weihua Zhuang
University of Waterloo
2 shared publications
Latest Publications
- 21 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Learning Technologies
- Vol. 19, 505-519
IAEAgent: LLM-Based Educational Agent for Instructional Alignment Evaluation in University Courses
In university courses, instructional alignment evaluation reflects the degree of alignment between the intended teaching requirements and the actual content delivered by teachers. It can not only provide university administrative departments with evidence about actual teaching performance, but also facilitate teachers' targeted reflection and continuous instructional improvement. However, there is still a lack of a theory-grounded, course-aware, and quantitatively interpretable instructional alignment evaluator. To address this issue, in this article, we propose a new evaluation paradigm—an LLM-based educational agent (IAEAgent)—for instructional alignment evaluation in university courses. Firstly, by injecting teaching materials such as textbooks using parameter-efficient fine-tuning (PEFT), a course-specific LLM with in-depth disciplinary comprehension is constructed. Then, based on this LLM and employing prompt engineering, the intended and actual instructional knowledge graphs (KGs) can be respectively extracted from teaching requirement documents and teachers' actual audio transcripts of a lesson. Subsequently, guided by teaching quality evaluation theories, a quantitative analysis of instructional alignment is conducted for these two KGs across three crucial dimensions: knowledge coverage completeness, knowledge depth consistency, and knowledge structure coherence. Finally, based on the obtained analytical results, a teaching quality evaluation report containing teaching diagnostics and improvement suggestions is generated. Extensive experiments demonstrate that the IAEAgent excels in comprehensive and fine-grained instructional alignment evaluation. Furthermore, ablation studies validate the effectiveness of each step within the IAEAgent.
- 21 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Learning Technologies
- Vol. 19, 505-519
IAEAgent: LLM-Based Educational Agent for Instructional Alignment Evaluation in University Courses
In university courses, instructional alignment evaluation reflects the degree of alignment between the intended teaching requirements and the actual content delivered by teachers. It can not only provide university administrative departments with evidence about actual teaching performance, but also facilitate teachers' targeted reflection and continuous instructional improvement. However, there is still a lack of a theory-grounded, course-aware, and quantitatively interpretable instructional alignment evaluator. To address this issue, in this article, we propose a new evaluation paradigm—an LLM-based educational agent (IAEAgent)—for instructional alignment evaluation in university courses. Firstly, by injecting teaching materials such as textbooks using parameter-efficient fine-tuning (PEFT), a course-specific LLM with in-depth disciplinary comprehension is constructed. Then, based on this LLM and employing prompt engineering, the intended and actual instructional knowledge graphs (KGs) can be respectively extracted from teaching requirement documents and teachers' actual audio transcripts of a lesson. Subsequently, guided by teaching quality evaluation theories, a quantitative analysis of instructional alignment is conducted for these two KGs across three crucial dimensions: knowledge coverage completeness, knowledge depth consistency, and knowledge structure coherence. Finally, based on the obtained analytical results, a teaching quality evaluation report containing teaching diagnostics and improvement suggestions is generated. Extensive experiments demonstrate that the IAEAgent excels in comprehensive and fine-grained instructional alignment evaluation. Furthermore, ablation studies validate the effectiveness of each step within the IAEAgent.
- 27 April 2026
- Published by Springer Nature in Nature Sensors
- Vol. 1 (5) , 389-402
Advanced sensors for immersive multisensory communication
- 27 April 2026
- Published by Springer Nature in Nature Sensors
- Vol. 1 (5) , 389-402
Advanced sensors for immersive multisensory communication
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 1275-1279
Task-Oriented Video Compression via XAI-Guided Frame Filtering
Task-oriented video compression aims to eliminate redundancy while preserving task-critical information. However, existing spatial domain methods incur high computational overhead, whereas temporal domain approaches often rely on black-box confidence scores that may discard taskrelevant frames. These methods follow an implicit informationpreservation paradigm that entangles redundancy and relevance, yielding opaque decision-making. In this paper, we therefore reformulate it as an explicit and explainable frame filtering method. Instead of relying on black-box confidence scores, task relevance is quantified by an explainable saliency score grounded in explicit model attribution and statistical aggregation, providing a transparent criterion for frame filtering. This offline-defined saliency metric supervises a lightweight online regressor, so that each online decision inherits the same explainable semantic meaning while avoiding heavy computation at the edge. Specifically, frame filtering is decomposed into two sequential and explainable steps: a similarity detector first removes structurally redundant frames, and a saliency-guided filter then discards frames with limited contribution to the downstream task. Experimental results demonstrate the effectiveness of the proposed method.
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 1275-1279
Task-Oriented Video Compression via XAI-Guided Frame Filtering
Task-oriented video compression aims to eliminate redundancy while preserving task-critical information. However, existing spatial domain methods incur high computational overhead, whereas temporal domain approaches often rely on black-box confidence scores that may discard taskrelevant frames. These methods follow an implicit informationpreservation paradigm that entangles redundancy and relevance, yielding opaque decision-making. In this paper, we therefore reformulate it as an explicit and explainable frame filtering method. Instead of relying on black-box confidence scores, task relevance is quantified by an explainable saliency score grounded in explicit model attribution and statistical aggregation, providing a transparent criterion for frame filtering. This offline-defined saliency metric supervises a lightweight online regressor, so that each online decision inherits the same explainable semantic meaning while avoiding heavy computation at the edge. Specifically, frame filtering is decomposed into two sequential and explainable steps: a similarity detector first removes structurally redundant frames, and a saliency-guided filter then discards frames with limited contribution to the downstream task. Experimental results demonstrate the effectiveness of the proposed method.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- p. 1-9
Cognitive-Inspired Cross-Modal Communications
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- p. 1-9
Cognitive-Inspired Cross-Modal Communications
- 29 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73 (11) , 11432-11447
Cross-Modal Semantic Transmission Strategy for Mobile Scenarios
To fulfill the demands of emerging multi-modal services, the cross-modal semantic communication paradigm comes into being. It fully utilizes potential semantic correlations among modalities to address polysemy and ambiguity issues, enhancing transmission reliability. However, applying cross-modal semantic communication in resource-constrained mobile scenarios introduces new challenges, including radio spectrum bandwidth limitations and fluctuations for the transmitter, and computing resource constraints for the receiver, which leads to potential transmission failures. To bridge this gap, this paper proposes a cross-modal semantic transmission strategy for mobile scenarios (MobileCMST). We first construct the framework for MobileCMST. Within this framework, a semantic encoder is designed to achieve redundancy elimination for visual and haptic signals. Then, a semantic delivery approach is developed to cope with bandwidth fluctuations and multipath fading channels. Finally, an efficient semantic decoder based on a visual-haptic semantic-integrated diffusion model is proposed. It employs the Mamba backbone to reconstruct high-quality signals with lightweight computational complexity. Extensive experiments demonstrate the excellent performance of the proposed MobileCMST strategy in resource-constrained mobile scenarios.
- 29 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73 (11) , 11432-11447
Cross-Modal Semantic Transmission Strategy for Mobile Scenarios
To fulfill the demands of emerging multi-modal services, the cross-modal semantic communication paradigm comes into being. It fully utilizes potential semantic correlations among modalities to address polysemy and ambiguity issues, enhancing transmission reliability. However, applying cross-modal semantic communication in resource-constrained mobile scenarios introduces new challenges, including radio spectrum bandwidth limitations and fluctuations for the transmitter, and computing resource constraints for the receiver, which leads to potential transmission failures. To bridge this gap, this paper proposes a cross-modal semantic transmission strategy for mobile scenarios (MobileCMST). We first construct the framework for MobileCMST. Within this framework, a semantic encoder is designed to achieve redundancy elimination for visual and haptic signals. Then, a semantic delivery approach is developed to cope with bandwidth fluctuations and multipath fading channels. Finally, an efficient semantic decoder based on a visual-haptic semantic-integrated diffusion model is proposed. It employs the Mamba backbone to reconstruct high-quality signals with lightweight computational complexity. Extensive experiments demonstrate the excellent performance of the proposed MobileCMST strategy in resource-constrained mobile scenarios.
- Top Cited PapersOpen Access
- 24 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-17
A wearable obstacle avoidance device for visually impaired individuals with cross-modal learning
It is challenging for wearable obstacle avoidance devices to simultaneously meet practical demands of high reliability, rapid response, long-lasting duration, and usable design. Here we report a wearable obstacle avoidance device, comprising a set of self-developed glasses (weighing ~400 grams, including an ~80 grams battery) and a common smartphone. Specifically, the glasses collect the multi-modal data for comprehensive environmental perception, including video and depth modalities, and implement a depth-aided video compression module. This module not only adaptively compresses video data to reduce transmission delay to the smartphone, but also operates on a customized FPGA board featuring a multi float-point vector unit streaming processing architecture, thereby facilitating responsive and energy-efficient obstacle detection. Additionally, we design a cross-modal obstacle detection module on the smartphone, which ensures reliable detection and provides user-friendly auditory and tactile alerts by utilizing cross-modal learning based on modal correlations. Multiple indoor and outdoor experimental results demonstrate 100% collision avoidance rates, delay of less than 320 ms, and duration of approximately 11 hours.
- 24 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-17
Top Cited PapersOpen AccessA wearable obstacle avoidance device for visually impaired individuals with cross-modal learning
It is challenging for wearable obstacle avoidance devices to simultaneously meet practical demands of high reliability, rapid response, long-lasting duration, and usable design. Here we report a wearable obstacle avoidance device, comprising a set of self-developed glasses (weighing ~400 grams, including an ~80 grams battery) and a common smartphone. Specifically, the glasses collect the multi-modal data for comprehensive environmental perception, including video and depth modalities, and implement a depth-aided video compression module. This module not only adaptively compresses video data to reduce transmission delay to the smartphone, but also operates on a customized FPGA board featuring a multi float-point vector unit streaming processing architecture, thereby facilitating responsive and energy-efficient obstacle detection. Additionally, we design a cross-modal obstacle detection module on the smartphone, which ensures reliable detection and provides user-friendly auditory and tactile alerts by utilizing cross-modal learning based on modal correlations. Multiple indoor and outdoor experimental results demonstrate 100% collision avoidance rates, delay of less than 320 ms, and duration of approximately 11 hours.
- 14 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 1650-1654
Cross-Attention and Cycle-Consistency-Based Haptic to Image Inpainting
With the rapid advancement of deep learning and multimedia technologies, image inpainting has made significant progress in generating desirable content for damaged images. To fill in the corrupted part of the image, existing methods either utilize intrinsic information within the visual modality for single-modal image inpainting or consider semantic correlations from non-visual modalities for cross-modal image inpainting. However, when the corrupted part is extensive, intrinsic information alone within visual modality is insufficient to inpaint the entire image. Cross-modal image inpainting methods cannot guarantee image quality, as they do not explore the relationship among the corrupted part of the image, the remaining part of the image, and the non-visual modality. In order to handle this issue, a haptic to image inpainting scheme is proposed. Specifically, feature extraction and cross-attention-based feature index are firstly constructed to find the correspondence between the corrupted part of the image and the haptic modality. Then, cycle-consistency-based feature translation is implemented to explore the intrinsic correlations among the above two modalities, facilitating the formation of the corrupted part. Finally, the formed corrupted part is combined with the remaining part to realize image inpainting. Experimental results show the effectiveness of the proposed scheme.
- 14 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 1650-1654
Cross-Attention and Cycle-Consistency-Based Haptic to Image Inpainting
With the rapid advancement of deep learning and multimedia technologies, image inpainting has made significant progress in generating desirable content for damaged images. To fill in the corrupted part of the image, existing methods either utilize intrinsic information within the visual modality for single-modal image inpainting or consider semantic correlations from non-visual modalities for cross-modal image inpainting. However, when the corrupted part is extensive, intrinsic information alone within visual modality is insufficient to inpaint the entire image. Cross-modal image inpainting methods cannot guarantee image quality, as they do not explore the relationship among the corrupted part of the image, the remaining part of the image, and the non-visual modality. In order to handle this issue, a haptic to image inpainting scheme is proposed. Specifically, feature extraction and cross-attention-based feature index are firstly constructed to find the correspondence between the corrupted part of the image and the haptic modality. Then, cycle-consistency-based feature translation is implemented to explore the intrinsic correlations among the above two modalities, facilitating the formation of the corrupted part. Finally, the formed corrupted part is combined with the remaining part to realize image inpainting. Experimental results show the effectiveness of the proposed scheme.
- 23 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 72 (10) , 6059-6072
Toward Generic Cross-Modal Transmission Strategy
Multi-modal services, integrating various modalities such as audio, visual, and haptic, have emerged as leading multimedia applications in the 5G era and beyond. To fulfill the demands for low latency, high reliability, and large capacity, cross-modal transmission schemes have been proposed. Typically, these schemes emphasize on either audio-visual or haptic modality, and prioritize flawless transmission of one modality to assist the other modality streaming. However, these prerequisite and assumption do not hold for generic multi-modal services and communication environments, where determining the priority of modality and guaranteeing flawless transmission becomes challenging. To address this fundamental problem, in this paper, we introduce a strategy toward generic cross-modal transmission, enabling visual and haptic modalities to assist each other as needed. The strategy includes a visual-haptic mutual stream delivery mechanism at the sender and a visual-haptic mutual signal reconstruction approach at the receiver. The former aims to eliminate redundancy in visual and haptic streams through mutual assistance, while the latter adaptively handles impaired, missing, or delayed visual or haptic signals by leveraging modality-aware knowledge transfer and semantic-aware signal generation techniques. The proposed strategy demonstrates excellent performance through experiments conducted on a standard multi-modal dataset and a practical visual-haptic communication platform.
- 23 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 72 (10) , 6059-6072
Toward Generic Cross-Modal Transmission Strategy
Multi-modal services, integrating various modalities such as audio, visual, and haptic, have emerged as leading multimedia applications in the 5G era and beyond. To fulfill the demands for low latency, high reliability, and large capacity, cross-modal transmission schemes have been proposed. Typically, these schemes emphasize on either audio-visual or haptic modality, and prioritize flawless transmission of one modality to assist the other modality streaming. However, these prerequisite and assumption do not hold for generic multi-modal services and communication environments, where determining the priority of modality and guaranteeing flawless transmission becomes challenging. To address this fundamental problem, in this paper, we introduce a strategy toward generic cross-modal transmission, enabling visual and haptic modalities to assist each other as needed. The strategy includes a visual-haptic mutual stream delivery mechanism at the sender and a visual-haptic mutual signal reconstruction approach at the receiver. The former aims to eliminate redundancy in visual and haptic streams through mutual assistance, while the latter adaptively handles impaired, missing, or delayed visual or haptic signals by leveraging modality-aware knowledge transfer and semantic-aware signal generation techniques. The proposed strategy demonstrates excellent performance through experiments conducted on a standard multi-modal dataset and a practical visual-haptic communication platform.