I will develop ai based object detection and tracking solutions

CORE DETECTION EXPERTISE:

• Foundation Models: Zero-shot detection, VLMs(Florence/paligemma 2) LoRA fine-tuning, CLIP/DINO optimization
• CNN Architectures: YOLO/RetinaNet (single-stage), R-CNN/Mask R-CNN (two-stage), FPN architectures
• Transformers: Swin/Swift Transformer, DETR, DINOv2 backbone integration
• Few-shot Learning: Transfer learning, knowledge distillation, active learning pipelines

ADVANCED TRACKING:

• SOTA MOT: ByteTrack, DeepSORT (custom embeddings), StrongSORT, Norfair
• Capabilities: Multi-camera re-ID, occlusion handling, trajectory forecasting

TECHNICAL STACK:

• Frameworks: MMDetection, Detectron2, TF-OD API, custom architecture
• Libraries: PyTorch, TensorFlow/Keras, OpenCV, Supervision

OPTIMIZATION:

• High-Performance: Triton Server, TensorRT, OpenVINO, DeepStream SDK
• Edge Deployment: Jetson optimization, quantization-aware training
• Data Management: ElasticSearch/MongoDB for metadata; ChromaDB for vectors

DOMAIN EXPERTISE:

• Industrial: Anomaly detection, PPE compliance, quality control
• Surveillance: Real-time threat monitoring, perimeter protection, behavior analysis
• Transportation: Vehicle tracking, ALPR, traffic analysis
• Retail: product recognition, automated inventory