- Develop an AI-powered smart surveillance camera system capable of scaling up to thousands of cameras
- Develop the HydraNets Architecture for Multi-Object Detection
- Deploy a face recognition system on the Jetson Nano
- Customize YOLO, RetinaFace, and other suitable models for optimal integration within the HydraNets architecture
- Manage the collection and labeling of real-world data for model training
- Deploy a license plate recognition model for application in traffic camera systems
- Optimize YOLO, RetinaFace, ArcFace, and other models for inference on TensorRT using FP32, FP16, and INT8 precisions
- Implement object detection (using models like YOLO, RetinaFace, and ArcFace), license plate recognition (DLC model), and other relevant functionalities on the QCS605 board (Qualcomm edge devices) using C++