AI Paper Atlas

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

HCL-FF: Hierarchical and Contrastive Learning for Forward-Forward Algorithm

Jie-En Yao, Hong-En Chen, C.-C. Jay Kuo

CVPR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 0 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 contrastive objective (V2 to V3) (ablation)HCL-FF · CIFAR-100accuracy: 54.65 -> 63.77Improved
add contrastive objective (V2 to V3) (ablation)HCL-FF · CIFAR-10accuracy: 88.80 -> 91.05Improved
add hierarchical objective (V2 to V4) (ablation)HCL-FF · CIFAR-100accuracy: 54.65 -> 61.61Improved
add hierarchical objective (V2 to V4) (ablation)HCL-FF · CIFAR-10accuracy: 88.80 -> 88.20Hurt
add strong data augmentation without hierarchy (V1 to V2) (ablation)HCL-FF baseline · CIFAR-10accuracy: 86.22 -> 88.80Improved
add strong augmentation with hierarchy and contrastive (V5 to V6) (ablation)HCL-FF · CIFAR-100accuracy: 65.22 -> 70.76Improved
add residual shortcuts (ablation)HCL-FF · CIFAR-100accuracy: 65.43 -> 70.76Improved
add residual shortcuts (ablation)HCL-FF · CIFAR-10accuracy: 90.85 -> 91.83Improved
use WordNet hierarchy instead of data-driven (ablation)HCL-FF · CIFAR-100accuracy: 70.76 -> 71.01No clear effect
use Word2Vec hierarchy instead of data-driven (ablation)HCL-FF · CIFAR-100accuracy: 70.76 -> 69.59Hurt
use balanced mapping instead of decremental (ablation)HCL-FF · CIFAR-100accuracy: 68.40 -> 70.76Improved
use selective interval prediction instead of last layer (ablation)HCL-FF · CIFAR-100accuracy: 67.42 -> 70.76Improved

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