AI Paper Atlas

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

Probing Sentiment-Oriented Pre-Training Inspired by Human Sentiment Perception Mechanism

Tinglei Feng, Jiaxuan Liu, Jufeng Yang

CVPR 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 11 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 sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet50 · FIAcc: 0.670 -> 0.707Improved
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet101 · UnBiasedEmoAcc: 0.823 -> 0.859Improved
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet18 · EmoticHamming: 0.163 -> 0.145Improved
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet50 · EmoticMacroF1: 0.041 -> 0.043Improved
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationVgg19 · Emotion6Hamming: 0.253 -> 0.171Improved
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet101 · Emotion6MacroF1: 0.825 -> 0.827No clear effect
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet50 · Emotion6KL: 0.403 -> 0.402Hurt
Use sentiment-oriented pre-training and amalgamation instead of ImageNet initializationResNet50 · AbstractKL: 0.568 -> 0.517Improved
Use sentiment-oriented pre-training rather than SimCLRFI models · FIAccuracy: 0.627 -> 0.707Improved
Add ST, HO, and HP feature regularization (ablation)ResNet50 amalgamation · FIAcc: 66.95 -> 68.92Improved
Add logits regularization from the ST model (ablation)ResNet50 amalgamation · FIAcc: 66.95 -> 66.38Hurt
Add logits regularization from the HP model (ablation)ResNet50 amalgamation · FIAcc: 66.95 -> 70.20Improved

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