Xin Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Xin Li has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Euee Seon Jang
Department of Computer Science, Hanyang University, Seoul, South Korea
2 shared publications Gwangsoon Lee
2 shared publicationsKwan-Jung Oh
2 shared publicationsEuee S. Jang
1 shared publication- Eun-Young Chang
Development of Core Technologies for Autonomous Searching Drones, Electronics and Telecommunications Research Institute (ETRI), Daejeon, Republic of Korea
1 shared publication Jaiyoung Oh
1 shared publication
Latest Publications
- Open Access
- 30 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 186785-186794
Separated Background and Foreground Coding for MPEG Immersive Video Coding
In this paper, we propose a new coding method based on the separation of background and foreground data for the MPEG Immersive Video (MIV) coding standard. The main contribution of this work is the development of a background and foreground data separation framework for MIV, along with the integration of spatio-temporal merge techniques to enhance background data reuse. Background data are generally more static than foreground data are. By leveraging the relatively static nature of background data, we enable its reuse by applying temporal merge across neighboring frames or spatial merge to convert multi view content into a single view. To extend the functionality of the current MIV, we implemented and made the separated background and foreground coding available and provided the necessary configuration to enable the proposed method. By extending the functionality, it is now possible to encode externally separated foreground and background data using the MIV encoder and decoder. Furthermore, the foreground and background bitstreams can be rendered in one scene or view by applying a spatio-temporal merge to the background. This technique can reduce the bitrate by approximately 5% and the time complexity by approximately 56% compared to the anchor. The proposed method was implemented and adopted in the second edition of the MIV coding standard for the Moving Picture Experts Group (MPEG).
- 30 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 186785-186794
Open AccessSeparated Background and Foreground Coding for MPEG Immersive Video Coding
In this paper, we propose a new coding method based on the separation of background and foreground data for the MPEG Immersive Video (MIV) coding standard. The main contribution of this work is the development of a background and foreground data separation framework for MIV, along with the integration of spatio-temporal merge techniques to enhance background data reuse. Background data are generally more static than foreground data are. By leveraging the relatively static nature of background data, we enable its reuse by applying temporal merge across neighboring frames or spatial merge to convert multi view content into a single view. To extend the functionality of the current MIV, we implemented and made the separated background and foreground coding available and provided the necessary configuration to enable the proposed method. By extending the functionality, it is now possible to encode externally separated foreground and background data using the MIV encoder and decoder. Furthermore, the foreground and background bitstreams can be rendered in one scene or view by applying a spatio-temporal merge to the background. This technique can reduce the bitrate by approximately 5% and the time complexity by approximately 56% compared to the anchor. The proposed method was implemented and adopted in the second edition of the MIV coding standard for the Moving Picture Experts Group (MPEG).
- Open Access
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 96135-96146
Efficient Inter-Prediction Method Using Reference Frame Accumulation for MPEG G-PCC
This paper presents a novel hybrid reference frame generation method to enhance the interprediction performance of geometry-based point cloud compression (G-PCC), a recent point cloud coding standard developed by the moving picture experts group (MPEG) that currently employs a single past reference frame for prediction. However, this single-reference approach may not fully capture the temporal and spatial correlation between frames, potentially limiting prediction performance. To address this limitation, we propose a hybrid mode selection scheme that chooses the suitable reference frame from a set of candidate frames, thereby leveraging the spatial and temporal correlation among consecutive past frames more effectively and improving coding efficiency. We introduce and evaluate three coding modes for the hybrid reference frame concept within the G-PCC test model: (1) entropy estimation-based mode selection, which selects the reference frame(s) that minimize the estimated entropy during inter-frame coding; (2) translation-based mode selection, which selects the reference frame(s) requiring the least translation to align with the current frame; and (3) lightweight entropy estimation-based mode selection, which minimizes entropy under a predefined translation constraint. Experimental results demonstrate that all proposed techniques outperform the current G-PCC standard: entropy estimation-based mode selection achieved average improvements of 0.86% in geometric coding efficiency and 0.81% overall. Translation-based mode selection improved these metrics by 0.59% and 0.56%, respectively, and the lightweight entropy estimation approach consistently yielded a 0.7% gain. In specific driving scenarios, all the proposed techniques showed significant improvements, reaching approximately 2.62% to 2.86% in geometric coding efficiency.
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 96135-96146
Open AccessEfficient Inter-Prediction Method Using Reference Frame Accumulation for MPEG G-PCC
This paper presents a novel hybrid reference frame generation method to enhance the interprediction performance of geometry-based point cloud compression (G-PCC), a recent point cloud coding standard developed by the moving picture experts group (MPEG) that currently employs a single past reference frame for prediction. However, this single-reference approach may not fully capture the temporal and spatial correlation between frames, potentially limiting prediction performance. To address this limitation, we propose a hybrid mode selection scheme that chooses the suitable reference frame from a set of candidate frames, thereby leveraging the spatial and temporal correlation among consecutive past frames more effectively and improving coding efficiency. We introduce and evaluate three coding modes for the hybrid reference frame concept within the G-PCC test model: (1) entropy estimation-based mode selection, which selects the reference frame(s) that minimize the estimated entropy during inter-frame coding; (2) translation-based mode selection, which selects the reference frame(s) requiring the least translation to align with the current frame; and (3) lightweight entropy estimation-based mode selection, which minimizes entropy under a predefined translation constraint. Experimental results demonstrate that all proposed techniques outperform the current G-PCC standard: entropy estimation-based mode selection achieved average improvements of 0.86% in geometric coding efficiency and 0.81% overall. Translation-based mode selection improved these metrics by 0.59% and 0.56%, respectively, and the lightweight entropy estimation approach consistently yielded a 0.7% gain. In specific driving scenarios, all the proposed techniques showed significant improvements, reaching approximately 2.62% to 2.86% in geometric coding efficiency.