FOF: A Fine-Grained Object Detection and Feature Extraction end-to-end Network

Currently, widely-used object detection can predict targets present in the training set. However, in fine-grained object detection tasks, such as commodity detection, the introduction of a new target class requires retraining the model, which significantly reduces the flexibility of the algorithm in applications. In response to this problem, we propose an end-to-end fine-grained object detection and feature extraction network(FOF). To detect and identify objects beyond the target category of the training set, the category output in the network Head is removed and replaced with a 128-dimensional feature vector. We used the ArcFace loss function to improve feature classification during training. Since there is no category output, an improved Non-Maximum Suppression(NMS) algorithm, Non-Maximum Suppression-Feature Similarity(NMS-FS), is proposed to distinguish same class and dissimilar class prediction boxes by feature similarity. During the inference, FOF outputs prediction boxes and feature vectors, and matches them with the feature vectors in the feature gallery to determine the detected object category and complete object detection and recognition. Experimental results indicate that FOF achieved high accuracy in both the MS COCO dataset and a large-scale and fine-grained Retail Product Checkout dataset(RPC). In addition, the method exhibits a low Equal Error Rate(EER) when identifying new categories, achieving the objective of detecting and identifying new categories without the need to retrain the model.