Diverse and Tailored Image Generation for Zero-shot Multi-label Classification

arXiv:2404.03144v1 Announce Type: new
Abstract: Recently, zero-shot multi-label classification has garnered considerable attention for its capacity to operate predictions on unseen labels without human annotations. Nevertheless, prevailing approaches often use seen classes as imperfect proxies for unseen ones, resulting in suboptimal performance. Drawing inspiration from the success of text-to-image generation models in producing realistic images, we propose an innovative solution: generating synthetic data to construct a training set explicitly tailored for proxyless training on unseen labels. Our approach introduces a …

abstract annotations arxiv attention capacity classification cs.cv diverse human image image generation image generation models inspiration labels performance predictions proxies success text text-to-image type zero-shot