Improving Multi-label Recognition using Class Co-Occurrence Probabilities

arXiv:2404.16193v1 Announce Type: cross
Abstract: Multi-label Recognition (MLR) involves the identification of multiple objects within an image. To address the additional complexity of this problem, recent works have leveraged information from vision-language models (VLMs) trained on large text-images datasets for the task. These methods learn an independent classifier for each object (class), overlooking correlations in their occurrences. Such co-occurrences can be captured from the training data as conditional probabilities between a pair of classes. We propose a framework to extend …

abstract arxiv class classifier complexity cs.ai cs.cv cs.lg cs.mm datasets eess.iv identification image images improving independent information language language models learn multiple object objects recognition text type vision vision-language vision-language models vlms