CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification

arXiv:2403.09281v1 Announce Type: new
Abstract: The CLIP (Contrastive Language-Image Pretraining) model has exhibited outstanding performance in recognition problems, such as zero-shot image classification and object detection. However, its ability to count remains understudied due to the inherent challenges of transforming counting--a regression task--into a recognition task. In this paper, we investigate CLIP's potential in counting, focusing specifically on estimating crowd sizes. Existing classification-based crowd-counting methods have encountered issues, including inappropriate discretization strategies, which impede the application of CLIP and result …

abstract arxiv challenges classification clip count cs.cv detection however image language object paper performance pretraining recognition regression through type zero-shot