Finally, we optimize the loss function to make it more suitable for the task of FSOD. Experimental results on PASCAL VOC and MS COCO demonstrate that our designed few-shot object detector has competitive performance with other state-of-the-art detectors....
TAL task alignment learning CNN convolutional neural network ROIs regions of interest FPN feature pyramid network GFL generalized focal loss GAP global average pooling GMP global max pooling BCE binary cross entropy CIOU center-weighted intersection over union FLOPS floating point operations CSP cross stag...
The CSP structure introduced in YOLO V5 can effectively strengthen the network feature extraction and fusion capabilities to achieve better performance. In addition, the PAN structure is derived from the idea of PANet in the field of object segmentation. Two feature maps of the same size are ...