Research map / Video / Video segmentation and tracking
Curriculum Learning With Infant Egocentric Videos
Saber Sheybani, Himanshu Hansaria, Justin Newell Wood, Linda B. Smith, Zoran Tiganj
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 |
|---|---|---|---|
| pretrain on youngest infant data (g0) in stage 1 instead of oldest (g2), random, or adult data | VideoMAE · SSv2-10, UCF101, ToyBox linear eval | linear evaluation score | Improved |
| youngest-age data yields larger last-layer gradient norm in stage 1 | VideoMAE · pretraining | encoder gradient norm | Improved |
| developmental order (g0-g1-g2) instead of anti-developmental order (g2-g1-g0) (ablation) | VideoMAE · SSv2-10, UCF101, ToyBox | linear evaluation score | Improved |
| developmental curriculum vs random and adult curricula (ablation) | VideoMAE · SSv2-10, UCF101, ToyBox | linear evaluation score | Improved |
| match spatial complexity (about 6% edge pixels) across age groups (ablation) | VideoMAE curricula · pretraining gradient norm and downstream | gradient norm and dev/anti-dev gap | Mixed |
| match temporal complexity by using single-frame samples (ablation) | VideoMAE curricula · UCF101, ToyBox | gradient norm and dev/anti-dev gap | Mixed |
| train on single frames (ablation) | VideoMAE · SSv2-10 | linear evaluation score | Hurt |
| developmental curriculum with JEPA-TT (ViT-Base) | JEPA-TT · ToyBox, CIFAR10 | linear evaluation score | Improved |
| developmental curriculum with SimCLR-TT (ResNet18) | SimCLR-TT · ToyBox, CIFAR10 | linear evaluation score | Improved |
| use a different optimizer (ablation) | VideoMAE curricula · ablation | curriculum effect pattern | No clear effect |
| use a different number of pretraining iterations (ablation) | VideoMAE curricula · ablation | curriculum effect pattern | No clear effect |
Closest papers by meaning
- Learning from an Infant’s Visual Experience (AAAI 2024 · similarity 0.87)
- Discovering Hidden Visual Concepts Beyond Linguistic Input in Infant Learning (CVPR 2025 · similarity 0.85)
- BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models (CVPR 2026 · similarity 0.84)
- BabyVision: Visual Reasoning Beyond Language (ICML 2026 · similarity 0.83)
- Alignment of CNN and Human Judgments of Geometric and Topological Concepts (AAAI 2025 · similarity 0.83)
- Object Concepts Emerge from Motion (NeurIPS 2025 · similarity 0.82)
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.
