CapDec: SOTA Zero Shot Image Captioning Using CLIP and GPT2, EMNLP 2022 (findings)
-
Updated
Jan 28, 2024 - Python
CapDec: SOTA Zero Shot Image Captioning Using CLIP and GPT2, EMNLP 2022 (findings)
kNN-Memory with ClipCap for Improved Long-Range Dependency Handling
An end-to-end pipeline that generates Farsi captions from images. It uses the ClipCap model for English captioning and Meta AI's SeamlessM4T v2 for translation, demonstrating a powerful combination of modern vision and language AI models.
To associate your repository with the clipcap topic, visit your repo's landing page and select "manage topics."