https://openai.com/research/clip

CLIP is a recent zero-shot model that can understand natural language and images simultaneously

https://openai.com/product/dall-e-2

DALL-E creates images from text and can perform zero-shot classification tasks as well.

Dear Titas De ,

Deep Learning has become an indispensable tool of Artificial Intelligence. Among its many applications, it is now commonly used to solve complex Computer Vision tasks through supervised learning.

https://www.v7labs.com/blog/zero-shot-learning-guide

Regards,

Shafagat

Thanks Christian Schmidt and Shafagat Mahmudova

Thanks Mirzakhmet Syzdykov

For zero-shot image classification, the siamese neural network can be used, however, CLIP, UNITER ((UNiversal Image-TExt Representation), ViLBERT (Vision-and-Language BERT), GPT (Generative Pre-trained Transformer) Models are SOTA for zero-shot image classification.

Thank you Hassan Sajjad . This is superb.