AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Temporal Knowledge Consistency for Unsupervised Visual Representation Learning

Weixin Feng, Yuanjiang Wang, Lihua Ma, Ye Yuan, Chi Zhang

ICCV 2021 · Label-efficient and robust learning › Knowledge distillation · approach: Few-shot and metric distillation · 7 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
add TKC temporal teachersMoCo v2 · ImageNet linear eval (200 epochs)top-1 accuracy: 67.5 -> 69.0Improved
add TKC temporal teachersBYOL · ImageNet linear eval (100 epochs)top-1 accuracy: 70.1 -> 72.4Improved
add TKC temporal teachersMoCo v2 (AlexNet) · ImageNet linear eval conv4top-1 accuracy: 39.0 -> 46.2Improved
add TKC temporal teachersMoCo v2 (AlexNet) · ImageNet linear eval conv5top-1 accuracy: 42.8 -> 44.0Improved
add one temporal teacher to EMA teacher (h=1 to h=2) (ablation)TKC · ImageNet linear eval (AlexNet)top-1 accuracy: 39.9 -> 42.2Improved
use three teachers instead of two (h=2 to h=3) (ablation)TKC · ImageNet linear eval (AlexNet)top-1 accuracy: 42.2 -> 43.5Improved
increase teachers beyond three (h=3 to h=6) (ablation)TKC · ImageNet linear eval (AlexNet)top-1 accuracy: 43.5 -> 41.8Hurt
use 4-layer MLP knowledge transformer instead of 2-layer (ablation)TKC · ImageNet linear eval (100 epochs)top-1 accuracy: 65.91 -> 66.21Improved
use bottleneck MLP knowledge transformer instead of 2-layer (ablation)TKC · ImageNet linear eval (100 epochs)top-1 accuracy: 65.91 -> 66.31Improved
add TKC with frozen backbone detectionMoCo v2 · PASCAL VOC detection (frozen backbone)AP: 51.7 -> 52.7Improved
add TKC with frozen backbone detectionMoCo v2 · COCO detection (frozen backbone)AP: 29.2 -> 34.7Improved
add TKC when fine-tuning on COCO detectionMoCo v2 · COCO detection (fine-tune)AP: 40.1 -> 40.4No clear effect

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