Deep Learning Researcher
1 week ago
Implement deep learning models for person detection, pose estimation, item classification, and action recognition.
Integrate latest research to test SOTA models on our custom datasets.
Devise models for aggregating visual and semantic features across multiple cameras perspectives and timescales.
Adapt pre-existing models to better handle the details of our particular problem set.
Experiment with model architectures, transfer learning strategies, data augmentation approaches, data sampling strategies, hyperparameter search.
Inspect model errors and formulate principled strategies to improve the accuracy of the system.