https://github.com/tootouch/active_learning-uncertainty_sampling
Pytorch implementation of "A New Active Labeling Method for Deep Learning"
https://github.com/tootouch/active_learning-uncertainty_sampling
Last synced: over 1 year ago
JSON representation
Pytorch implementation of "A New Active Labeling Method for Deep Learning"
- Host: GitHub
- URL: https://github.com/tootouch/active_learning-uncertainty_sampling
- Owner: TooTouch
- License: mit
- Created: 2023-03-05T13:23:40.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2023-03-22T02:31:43.000Z (over 3 years ago)
- Last Synced: 2025-01-11T20:44:28.794Z (over 1 year ago)
- Language: Python
- Size: 6.35 MB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Active Learning - Uncertainty Sampling
Pytorch implementation of "[A New Active Labeling Method for Deep Learning. IJCNN 2014](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6889457)".
# Environments
NVIDIA pytorch docker [ [link](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-22-12.html#rel-22-12) ]
```bash
docker pull nvcr.io/nvidia/pytorch:22.12-py3
```
`requirements.txt`
```bash
accelerate
wandb
torchvision
```
# Methods
`./query_strategies`
- Least Confidence
- Margin Sampling
- Entropy
# Results
**Experiment Setting**
- Model: ResNet18
- Batch Size: 128
- Optimizer: SGD
- Learning Rate: 0.1
- Learning Rate Scheduler: Cosine Annealing with Warm-up
**Active Learning**
- The Number of Initial Labeled Images: 5,000
- The Number of Query Images: 500
- The Number of Iteration: 20
# Reference
- cure-lab/deep-active-learning [ [link](https://github.com/cure-lab/deep-active-learning) ]
- Active Learning. Yi Zhang (CMU) [ [link](https://www.cs.cmu.edu/~tom/10701_sp11/recitations/Recitation_13.pdf) ]