Self-driving cars, drones hijacked by custom road signs : AI vision systems can be very literal readers

Your #AI driver is still not ready for prime time.

Hmmm, maybe this would be useful in #Ukraine.
#LVLMs #SelfDrivingCars #Drones

Award-Winning Project Explores Optical Document Recognition (ODR) in Princeton Prosody Archive Award-Winning Project Explores Optical Document Recognition (ODR) in Princeton Prosody Archive

Have a lot of character flaws—in your scanned documents, that is? CDH Senior Thesis Prize winner James Zhang ’25 can help! Read our Q&A with James on the limitations of #OCR, using #LVLMs to analyze visual elements, and building his open-source tool #MetaScribe. 🏆

FGAIF: Aligning Large Vision-Language Models with Fine-grained AI Feedback

Liqiang Jing, Xinya Du

Action editor: Hanwang Zhang

https://openreview.net/forum?id=Qhfw5CUVd7

#ai #vision #lvlms

Read the preprint 📜: arxiv.org/abs/2411.12174

Thanks to our amazing team for pushing this forward, and to ACL for recognizing this effort!

with @trilokpadhi.bsky.social, Rahul Garg, Hemang Jain, @pkprofgiri.bsky.social

#ACL2025 #NeuroSymbolicAI #CommonsenseAI #LVLMs #LLMs #SocialComputing

4/4 🧵

Our work develops a #NeurosymbolicAI approach, KID-VLM, that combines the #LVLMs and commonsense #KnowledgeGraphs to detect subtle and context-rich cues of #OnlineToxicity.

2/4 🧵