Problem
Research questions and friction points this paper is trying to address.
image compression
low-latency decoding
efficient entropy coding
resource-constrained hardware
PULSE: Unlocking Practical Image Compression on Single-Thread CPU
🤖 AI Summary
为解决资源受限硬件上的图像压缩计算成本高问题，PULSE通过低复杂度神经接收器、高效熵编码及人类-大模型协作优化架构的方法实现快速解码。
📝 Abstract
Despite recent progress in learned image compression, existing methods remain computationally expensive on resource-constrained hardware, particularly CPUs. We introduce PULSE, a practical codec that enables (1) low-latency decoding on diverse hardware platforms with an ultra-low-complexity 5.2 kMAC/pixel neural receiver, and (2) efficient bit-exact entropy coding with an integer linear CDF predictor and a meta prior. To recover compression performance under this tight budget, we introduce an agentic evolution process guided by heuristic probes that iteratively improves the architecture through human-LLM collaboration. PULSE decodes a 1080p image in 126 ms on a single CPU thread while achieving compression performance comparable to HM. After perceptual optimization, PULSE competes with larger perceptual codecs like MS-ILLM. Codes are at https://github.com/microsoft/GenCodec/tree/main/PULSE
Innovation
Methods, ideas, or system contributions that make the work stand out.
low-latency decoding
ultra-low-complexity neural receiver
integer linear CDF predictor
meta prior
agentic evolution process
🔎 Similar Papers
No similar papers found.
