Zero-shot VLMs already ate object detection. Group emotion recognition is next.
中文摘要
零样本VLM在无需训练和标签的情况下，正将能力从目标检测扩展至群体情绪识别，但模型评估仍具挑战。
English Summary
Zero-shot VLMs are applying prompt-based recognition to group emotions without training or labels, though accurate model evaluation remains a challenge.
Original Excerpt
No training, no labels, one prompt. The hard part wasn’t reading a crowd’s mood, it was scoring the models without fooling myself. Continue reading on Towards Deep Learning »