AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Learning Invariant Representations using Inverse Contrastive Loss

Aditya Kumar Akash, Vishnu Suresh Lokhande, Sathya N. Ravi, Vikas Singh

AAAI 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 5 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 ICL instead of the unregularized VAEVAE · MNISTadversarial invariance measure: 46 \pm 4 -> 32 \pm 0Improved
Use ICL instead of the strongest reconstruction baseline, CAIVAE · MNISTreconstruction error: 11.8 \pm 0.3 -> 16.6 \pm 0.1Hurt
Use ICL instead of MIVIB · Adultadversarial invariance measure: 78 \pm 2 -> 75 \pm 2Improved
Use ICL instead of the unregularized VIBVIB · Adultprediction accuracy: 84 \pm 0 -> 83 \pm 0Hurt
Use ICL instead of MMD_fVIB · Germanprediction accuracy: 74 \pm 1 -> 75 \pm 2Improved
Use ICL instead of MMD_{-s}VIB · Germanadversarial invariance measure: 75 \pm 2 -> 75 \pm 2No clear effect
Use ICL instead of UAIVIB · MNIST-ROTprediction accuracy: 98 -> 96 \pm 0Hurt
Use ICL instead of UAIVIB · MNIST-ROTadversarial invariance measure: 34 -> 33 \pm 1Improved
Use ICL instead of MMD_fVIB · ADNIadversarial invariance measure: 49 \pm 3 -> 46 \pm 7Improved
Use ICL instead of MMD_fVIB · ADNIprediction accuracy: 86 \pm 1 -> 84 \pm 3Hurt
Use ICL instead of CAIResNet-18 · Adult (Age)adversarial invariance measure (MSE): 129 \pm 10 -> 161 \pm 15Improved
Use ICL instead of UAIResNet-18 · Adult (Age)prediction accuracy: 84 \pm 0 -> 83 \pm 0Hurt

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