Research map / Architectures and efficiency / RNNs and normalization
Rep-Net: Efficient On-Device Learning via Feature Reprogramming
Li Yang, Adnan Siraj Rakin, Deliang Fan
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 Rep-Net instead of TinyTL | ProxylessNAS-Mobile · Aircraft | accuracy: 75.4 -> 77.4(↑2%) | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile@320 · CUB | accuracy: 81.0 -> 82.3(↑1.3%) | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · Flowers | accuracy: 95.5 -> 96.1 | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · Cars | accuracy: 85.0 -> 85.8 | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · CUB | accuracy: 77.1 -> 77.8 | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · Food | accuracy: 79.7 -> 80.5 | Improved |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · Pets | accuracy: 91.8 -> 91.8 | No clear effect |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · CIFAR10 | accuracy: 95.9 -> 95.9 | No clear effect |
| Use Rep-Net instead of TinyTL | ProxylessNAS-Mobile · CIFAR100 | accuracy: 81.4 -> 81.9 | Improved |
| Add dual feature exchange so both backbone and Rep-Net features are updated (ablation) | Rep-Net · CUB | accuracy: 65.5 -> 77.1 | Improved |
| Add backbone features to the connector exchange (ablation) | Rep-Net · CUB | accuracy: 67.8 -> 77.1 | Improved |
| Add Rep-Net features to the connector exchange (ablation) | Rep-Net · CUB | accuracy: 71.4 -> 77.1 | Improved |
Closest papers by meaning
- TinyTrain: Resource-Aware Task-Adaptive Sparse Training of DNNs at the Data-Scarce Edge (ICML 2024 · similarity 0.86)
- Beyond Low-rank Decomposition: A Shortcut Approach for Efficient On-Device Learning (ICML 2025 · similarity 0.84)
- TinyFoA: Memory Efficient Forward-Only Algorithm for On-Device Learning (AAAI 2025 · similarity 0.84)
- Model Reprogramming: Resource-Efficient Cross-Domain Machine Learning (AAAI 2024 · similarity 0.83)
- RECAST: Reparameterized, Compact weight Adaptation for Sequential Tasks (ICLR 2025 · similarity 0.83)
- TinyTTA: Efficient Test-time Adaptation via Early-exit Ensembles on Edge Devices (NeurIPS 2024 · 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.
