Research map / Label-efficient and robust learning / Self-supervised and contrastive
Information Competing Process for Learning Diversified Representations
Jie Hu, Rongrong Ji, ShengChuan Zhang, Xiaoshuai Sun, Qixiang Ye, Chia-Wen Lin, Qi Tian
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use ICP's two-part diversified representation | VGG16 · CIFAR-10 test | classification error (%): 6.67 -> 6.10 | Improved |
| use ICP's two-part diversified representation | GoogLeNet · CIFAR-10 test | classification error (%): 4.92 -> 4.26 | Improved |
| use ICP's two-part diversified representation | ResNet20 · CIFAR-100 test | classification error (%): 31.91 -> 28.13 | Improved |
| use ICP's two-part diversified representation | DenseNet40 · CIFAR-100 test | classification error (%): 27.55 -> 24.52 | Improved |
| combine information minimization and maximization constraints with competition | β-VAE · dSprites | MIG: 0.22 -> 0.48 | Improved |
| combine information minimization and maximization constraints with competition | β-VAE · 3D Faces | MIG: 0.54 -> 0.73 | Improved |
| add ICP reconstruction approach | β-VAE · CelebA | MSE: 9.2*10^{-3} -> 8.5*10^{-3} | Improved |
| add ICP reconstruction approach | β-VAE · CelebA | SSIM: 0.60 -> 0.62 | Improved |
| remove information-diversifying and competing constraints (ablation) | ICP · dSprites | MIG: 0.48 -> 0.33 | Hurt |
| remove information-diversifying and competing constraints (ablation) | ICP · 3D Faces | MIG: 0.73 -> 0.26 | Hurt |
| remove the information-competing objective (ablation) | ICP · dSprites | MIG: 0.48 -> 0.20 | Hurt |
| remove the information-competing objective (ablation) | ICP · 3D Faces | MIG: 0.73 -> 0.57 | Hurt |
Closest papers by meaning
- Competing Mutual Information Constraints with Stochastic Competition-Based Activations for Learning Diversified Representations (AAAI 2022 · similarity 0.87)
- I-Con: A Unifying Framework for Representation Learning (ICLR 2025 · similarity 0.84)
- On Mutual Information Maximization for Representation Learning (ICLR 2020 · similarity 0.82)
- Connecting Jensen–Shannon and Kullback–Leibler Divergences: A New Bound for Representation Learning (NeurIPS 2025 · similarity 0.81)
- Convolutional Channel-Wise Competitive Learning for the Forward-Forward Algorithm (AAAI 2024 · similarity 0.81)
- Diffusion Model with Cross Attention as an Inductive Bias for Disentanglement (NeurIPS 2024 · similarity 0.81)
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.
