---
格式版本: 2
标题: "Every Packet Counts: Dispersing Information for Loss-Resilient Learned Image Compression"
原文链接: "https://arxiv.org/abs/2608.11096"
发布日期: "2026-08-11"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Tue, 11 Aug 2026 16:01:03 UTC (2,718 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-12T11:18:39+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-12T11:16:29+08:00"
入库时间: "2026-08-12T03:18:39.426Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=performance&searchtype=all"
匹配关键词:
  - "performance"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 5
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "论文为计算机视觉领域，研究图像压缩，与超节点/AI Rack/机柜级AI基础设施完全无关。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-12T11:19:06+08:00"
AI主题相关性: 0
AI来源权威性: 5
AI新颖性: 0
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 0
采集批次: "2026年8月12日2点04分20秒"
采集批次ID: "20260812-020420-289"
去重键: "https://arxiv.org/abs/2608.11096"
---

## Computer Science > Computer Vision and Pattern Recognition

## Title:Every Packet Counts: Dispersing Information for Loss-Resilient Learned Image Compression

Authors:[Yuhang Wei](https://arxiv.org/search/cs?searchtype=author&query=Wei,+Y) (1), [Chuqin Zhou](https://arxiv.org/search/cs?searchtype=author&query=Zhou,+C) (1), [Yibo Shi](https://arxiv.org/search/cs?searchtype=author&query=Shi,+Y) (2), [Jing Wang](https://arxiv.org/search/cs?searchtype=author&query=Wang,+J) (2), [Guo Lu](https://arxiv.org/search/cs?searchtype=author&query=Lu,+G) (1) ((1) Shanghai Jiao Tong University, (2) Huawei Technologies Ltd.)

[View PDF](https://arxiv.org/pdf/2608.11096) [HTML (experimental)](https://arxiv.org/html/2608.11096v1)

> Abstract:Learned image compression (LIC) has achieved impressive rate-distortion performance. However, existing methods remain highly vulnerable to packet loss, a common challenge in satellite and emergency communications. This vulnerability stems from non-uniform information distribution at the packetization stage and sequential decoding dependencies at the entropy coding stage. We propose an end-to-end loss-resilient image compression scheme that addresses both. Before packetization, we introduce an Inter-Channel Redistribution (ICR) mechanism to redistribute channel energy, preventing critical information concentrating in a small subset of channels. Then, an Interleaved Channel Grouping (ICG) strategy partitions latent channels in a strided manner to disperse information across packets, with each packet kept within constrained sizes. To limit cascading errors from lost packets, we adopt a two-layer dual-branch autoregressive structure to shorten the dependency chain. Extensive experiments demonstrate that our method consistently outperforms existing approaches in both reconstruction quality and stability. At 20% packet loss, it achieves an average PSNR gain of 1.84 dB over LossResilientLIC while reducing PSNR variance by an order of magnitude. Notably, trained under uniform random loss only, our model generalizes to bursty loss modeled by the Gilbert-Elliott channel, outperforming methods explicitly trained for such conditions.

| Comments: |  |
| --- | --- |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| ACM classes: | I.4.2; I.4.10; I.2.6 |
| Cite as: | [arXiv:2608.11096](https://arxiv.org/abs/2608.11096) \[cs.CV\] |
|  | (or [arXiv:2608.11096v1](https://arxiv.org/abs/2608.11096v1) \[cs.CV\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.11096](https://doi.org/10.48550/arXiv.2608.11096) |
| Related DOI: | [https://doi.org/10.1145/3767308.3836205](https://doi.org/10.1145/3767308.3836205) |

## Submission history

From: Yuhang Wei \[[view email](https://arxiv.org/show-email/ef369e64/2608.11096)\]  
**\[v1\]** Tue, 11 Aug 2026 16:01:03 UTC (2,718 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.11096) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))