mlsanityData to 14 Oct 2026

PULSE: Unlocking Practical Image Compression on Single-Thread CPU

arXiv 2609.18602 ↗First posted 16 Sep 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

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

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 32 of its 41 references that are in the map. First posted on arXiv 16 Sep 2026.

Builds on

32 of 41 references.

References · 32

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  2. 2023Segment Anythingcited by 8,234
  3. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  4. 2022Scalable Diffusion Models with Transformerscited by 6,054
  5. 2017Squeeze-and-Excitation Networkscited by 4,789
  6. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,849
  7. 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,552
  8. 2016Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Networkcited by 1,204
  9. 2018Variational image compression with a scale hyperpriorcited by 935
  10. 2020Image Quality Assessment: Unifying Structure and Texture Similaritycited by 811
  11. 2016End-to-end Optimized Image Compressioncited by 696
  12. 2018Joint Autoregressive and Hierarchical Priors for Learned Image Compressioncited by 664
20 more
  1. 2020Learned Image Compression with Discretized Gaussian Mixture Likelihoods and Attention Modulescited by 590
  2. 2015An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognitioncited by 520
  3. 2020High-Fidelity Generative Image Compressioncited by 338
  4. 2022ELIC: Efficient Learned Image Compression with Unevenly Grouped Space-Channel Contextual Adaptive Codingcited by 301
  5. 2023Learned Image Compression with Mixed Transformer-CNN Architecturescited by 211
  6. 2022Hybrid Spatial-Temporal Entropy Modelling for Neural Video Compressioncited by 128
  7. 2023Improving Statistical Fidelity for Neural Image Compression with Implicit Local Likelihood Modelscited by 94
  8. 2022Multi-Realism Image Compression with a Conditional Generatorcited by 80
  9. 2025Towards Practical Real-Time Neural Video Compressioncited by 80
  10. 2023MLIC++: Linear Complexity Multi-Reference Entropy Modeling for Learned Image Compressioncited by 58
  11. 2023C3: High-performance and low-complexity neural compression from a single image or videocited by 55
  12. 2022COOL-CHIC: Coordinate-based Low Complexity Hierarchical Image Codeccited by 51
  13. 2023EVC: Towards Real-Time Neural Image Compression with Mask Decaycited by 49
  14. 2023Computationally-Efficient Neural Image Compression with Shallow Decoderscited by 23
  15. 2022Post-Training Quantization for Cross-Platform Learned Image Compressioncited by 14
  16. 2026CoD-Lite: Real-Time Diffusion-Based Generative Image Compressioncited by 7
  17. 2025Learning Switchable Priors for Neural Image Compressioncited by 5
  18. 2021End-to-end optimized image compression with competition of prior distributionscited by 4
  19. 2026What Matters in Practical Learned Image Compressioncited by 3
  20. 2026MLVC: Multi-platform Learned Video Codec for Real-World Deploymentcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Towards image compression with perfect realism at ultra-low bitrates
  2. 2026Rate-Adaptive One-Step Diffusion Compression for AIGC Images
  3. 2022Evaluating the Practicality of Learned Image Compression
  4. 2025Single-step Diffusion for Image Compression at Ultra-Low Bitrates
  5. 2026FlowCodec: One-Step Flow Prior for Generative Image Compression
  6. 2025PerCoV2: Ultra-Low Bit-Rate Perceptual Image Compression via Query-Based 1D Multimodal Image Tokens
  7. 2024Controllable Distortion-Perception Tradeoff Through Latent Diffusion for Neural Image Compression
  8. 2024Enhancing the Rate-Distortion-Perception Flexibility of Learned Image Codecs with Conditional Diffusion Decoders