Скачать или смотреть Learning Transferable Visual Models From Natural Language Supervision (CLIP) ft. Syeda Aiman Mumtaz

Join us for AI Paper Reading Clubs: Day 4, featuring Syeda Aiman Mumtaz, Women in Tech Lead at GDGOC CUI Islamabad.

In this session, we explore “Learning Transferable Visual Models From Natural Language Supervision (CLIP)”, a landmark paper that bridges computer vision and natural language by training models directly on image–text pairs. CLIP enables zero-shot learning, allowing models to generalize across tasks without any task-specific retraining.

If you’re interested in multimodal AI, foundation models, and zero-shot learning, this premiere is for you.

🎤 Featured Speaker: Syeda Aiman Mumtaz (Women in Tech Lead at GDGOC CUI)
Connect with Syeda on LinkedIn:   / syeda-aiman-mumtaz-sherazi-b32a0b280  

💡 What You'll Learn:
• Vision–language models using a shared embedding space for images and text
• Dual encoder architecture with ViT or ResNet for images and a Transformer for text
• Contrastive learning to align image–text pairs while separating thousands of negatives per batch
• Zero-shot classification using text prompts without any retraining
• Cross-modal retrieval and strong real-world performance across OCR, action recognition, and image retrieval

Slide Deck: https://drive.google.com/drive/folder...

Stay connected for more updates and future events on our Social Media:
•⁠ ⁠Commudle: https://www.commudle.com/communities/...
•⁠ Discord:   / discord  
•⁠ ⁠Facebook Group:   / tfugislamabad  
•⁠ ⁠Instagram:   / tfugisl  
•⁠ ⁠LinkedIn:   / tfugisl  
•⁠ Slack: https://join.slack.com/t/tfugislamaba...
•⁠ ⁠Twitter:   / tfugislamabad  
•⁠ ⁠WhatsApp Community: https://chat.whatsapp.com/HCAoCKBHahi...

Don’t miss out on more engaging AI Paper Reading Clubs and be part of our growing community!