AI Paper Atlas

Research map / Architectures and efficiency / Quantization

Video Coding for Machines: Compact Visual Representation Compression for Intelligent Collaborative Analytics

Wenhan Yang, Haofeng Huang, Yueyu Hu, Ling‐Yu Duan, Jiaying Liu

TPAMI 2024 · Architectures and efficiency › Quantization · approach: Post-training weight quantization · 91 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use codebook hyperprior instead of IDFCfeature compression · Semantic segmentationbit-rate (bpp): 0.026 -> 0.013Improved
use codebook hyperprior instead of plain hyperpriorfeature compression · Semantic segmentationmIoU: 25.42% -> 29.35%Improved
use codebook hyperprior instead of plain hyperpriorfeature compression · Scene classificationbit-rate (bpp): 0.0102 -> 0.0069Improved
use codebook hyperprior instead of IDFCfeature compression · Scene classificationtest accuracy: 65.43% -> 59.47%Hurt
use codebook hyperprior instead of IDFCfeature compression · Reshadingbit-rate (bpp): 0.0959 -> 0.0134Improved
compress tasks jointly in grouped Trinity setting instead of per-task Customized (ablation)multi-task compression · Six Taskonomy taskstotal bit-rate (bpp): 0.059 -> 0.049Improved
group six tasks in Trinity instead of non-optimal Trinity* (ablation)multi-task compression · Scene classificationaccuracy: 59.64% -> 71.08%Improved
compress all six tasks in one model (Hex) instead of Trinity (ablation)multi-task compression · Scene classificationaccuracy: 71.08% -> 62.18%Hurt
compress all six tasks in one model (Hex) instead of Trinity (ablation)multi-task compression · Semantic segmentationmIoU: 18.14% -> 20.30%Improved
add full pixel reconstruction to the Trinity group (Trinity+) (ablation)multi-task compression · Scene classificationaccuracy and bpp: 71.08% -> 64.62%; 0.016 -> 0.052Hurt
use BPG image codec instead of learned task featuresBPG-43 · Scene classificationaccuracy: 71.08% -> 50.05%Hurt
add source feature of unseen task to the compression (Source+) vs Internal (ablation)unseen task support · Object classificationaccuracy: 51.06% -> 53.50%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy