Working on fine-tuning YOLOv10 for video object detection using COCO-pretrained weights. Leveraging Kaggle GPUs for training and integrating trackers like ByteTrack/DeepSORT for stable multi-object tracking across frames. Focused on building robust, real-time detection models for diverse video scenarios.