F-VLM: open-vocabulary object detection upon frozen vision and language models

3 points| shantanu_sharma | 2 years ago |ai.googleblog.com | reply