SETI is an acronym for the Search for Extraterrestrial Intelligence. It is an effort to detect evidence of technological civilizations that may exist elsewhere in the universe. 7 type of SETI signals are present in the dataset:
- brightpixel
- narrowband
- narrowbanddrd
- noise
- squarepulsednarrowband
- squiggle
- squigglesquarepulsednarrowband
Our objective is to differenciate those signals properly. Dataset is available at Kaggle. This dataset is a smaller version of Machine Learning for the Search for Extraterrestrial Intelligence Hackathon & Code Challenge dataset.
| Model | Accuracy |
|---|---|
| KNN+ResNet101 | 30.71% |
| KNN+VGG16 | 52.38% |
| Model | Training epoch | Accuracy |
|---|---|---|
| VGG16 | 50 | 85.71% |
| EfficientNetV2S -fine tuned | 20 | 86.29% |
| Swin-T -fine tuned | 20 | 81% |