Deep Learning Approaches for Fashion and Clothing Detection: A Comprehensive Survey of Recent Advances
The intersection of computer vision and fashion e-commerce has witnessed remarkable progress with the advent of deep learning technologies. This comprehensive survey examines recent advances in fashion and clothing detection systems, synthesizing findings from five major research directions: video-to-shop retrieval, clothing classification using YOLO architectures, ensemble-based recommendation systems, fashion attribute detection, and outfit compatibility prediction. We analyze benchmark datasets including DeepFashion, DeepFashion2, Fashion-MNIST, and emerging Roboflow collections, while evaluating the performance of state-of-the-art architectures including YOLOv5, Faster R-CNN, EfficientNet, MobileNet, and transformer-based models. Our analysis reveals that single-stage detectors achieve superior inference speeds (up to 45 FPS) while maintaining competitive accuracy, ensemble methods improve classification robustness by 8-12%, and multi-task learning frameworks enhance retrieval prec...