Schedule

The workshop will be held on July 3, 2026 at the Guangzhou Yuexiu International Congress Center. See the venue on Google Maps

Room: 602B

Some accepted papers will be selected for spotlight presentations.

For details on the invited talks and the speakers, please see below.

Timetable

Chairs: Jun Chen, Ezgi Ozyilkan, Yong Fang, and Elza Erkip

Time (UTC-4) Event Invited Speaker / Spotlight Paper / Additional Notes
09:50 - 11:10 Technical Session 1  
09:50 – 10:30 Keynote Presentation 1 Aaron Wagner: Wasserstein Distortion: Unifying Fidelity and Realism
10:30 - 10:50 Oral Presentation 1 Changshuo Wang, Zijian Liang, Kai Niu, Ping Zhang: A Theoretical Framework for Rate-Distortion Limits in Learned Image Compression
10:50 - 11:10 Oral Presentation 2 Tong Wu, Zhiyong Chen, Guo Lu, Li Song, Feng Yang, Meixia Tao, Wenjun Zhang: Joint Source-Channel-Generation Coding: From Distortion-Oriented Reconstruction to Semantic-Consistent Generation
11:10 - 11:30 Tea Break  
11:30 - 12:50 Technical Session 2  
11:30 - 12:10 Keynote Presentation 2 Bo Bai: Forget BIT, It is all about TOKEN: Towards Semantic Information Theory for LLMs
12:10 - 12:50 Invited Talk 1 Yanjun Han: Next-symbol Prediction from Compression: Statistical and Computational Foundations
12:50 - 14:00 Lunch Break and Poster Session  
14:00 - 15:20 Technical Session 3  
14:00 - 14:40 Keynote Presentation 3 Zoe Liu: AV1 for Video AI: From Bit Compression to Token Compression
14:40 - 15:00 Oral Presentation 3 Chenghong Bian, Yulin Shao, Deniz Gündüz: Towards AI-Native Fronthaul: Neural Compression for NextG Cloud RAN
15:00 - 15:20 Oral Presentation 4 Gergely Flamich, Spencer Hill: Singular Relative Entropy Coding with Bits-Back Rejection Sampling
15:20 - 15:40 Tea Break  
15:40 - 18:00 Technical Session 4  
15:40 - 16:20 Invited Talk 2 Xueyan Niu: Compression for LLMs and LLMs for Compression
16:20 - 17:00 Invited Talk 3 Wenqi Shao: Towards Practical Foundation Model Compression by Quantization and Linear Attention
17:00 - 18:00 Panel and Open Discussion  

Accepted posters:

  1. MultiTok: Variable-Length Tokenization for Efficient LLMs Adapted from LZW Compression by Noel Elias; Homa Esfahanizadeh; Kaan Kale; Sriram Vishwanath; Muriel Médard
  2. SCKM: Symmetric Co-Skew Moment for User Selection in Federated Learning by Liangyan Li; Yangyi Liu; Yimo Ning; Stefano Rini; Jun Chen
  3. A Distortion-Aware Discrete Semantic Communication Framework for Multiple-Access Channels by Yiqiao Zhou; Na Wang; Xiaolong Hong; Kangning Ma; Shuo Shao
  4. Nonsquare 2D Orthogonal Matching Pursuit with Karhunen-Loeve Transform and Trellis Coded Quantization by Linlong Guo; Yize Yang; Yong Fang
  5. Serial Concatenation of Compander and Trellis Coded Quantizer by Linlong Guo; Yize Yang; Yong Fang
  6. Hamming Distance Spectrum-Based Dynamic Pruning Method in Distributed Arithmetic Coding by Junwei Zhou; Haoyun Xiao; Jianwen Xiang; Yanchao Yang
  7. Rate-Distortion-Complexity: Bridging Shannon Information and Neural Network Expressivity by Jiancheng Tang; Qianqian Yang
  8. The Entropy of Early Language: Benchmarking Source Coding on the CHILDES Corpus by Erick Oduniyi