AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Distributed Composite Quantization

Weixiang Hong, Jingjing Meng, Junsong Yuan

AAAI 2018 · Architectures and efficiency › RNNs and normalization · approach: Dataset distillation and data-efficient training · 10 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
distributed training (DCQ) instead of centralized CQComposite Quantization · Holidays Fisher 64-bitmAP: 0.560 -> 0.554No clear effect
distributed training (DCQ) instead of centralized CQComposite Quantization · SIFT1M 128-bitrecall@1 (%): 54.43 -> 53.27 (4-node tree)Hurt
line topology instead of binary tree topology (ablation)DCQ · SIFT1M 128-bititerations to converge: 22 -> 26 (4 nodes)Hurt
line topology instead of binary tree topology (ablation)DCQ · SIFT1M 128-bitrecall@1 (%): 53.27 -> 52.78 (4 nodes); 52.19 -> 52.55 (16 nodes)No clear effect
increase number of nodes from 1 to 16 (tree) (ablation)DCQ · SIFT1M 128-bitrecall@1 (%): 54.43 -> 52.19No clear effect
distributed (16-node tree) versus single-machine trainingCQ · SIFT1M 128-bittraining time (min): 328 -> about 63Improved
addition-based (CQ/DCQ) instead of partition-based quantization (PQ)quantizer · UKBench Fisher 64-bitscore: 1.637 -> 2.315Improved

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