Welcome fellow researchers to the DeCLaRe Lab! We are a small group of researchers working on challenging problems in AI. To know more about our work, please browse our catalog of publications.

Here are some key research by our lab that got some attention:

• Multimodal Representation Learning/Multimodal Fusion: One of the first works to show multimodal fusion can be achieved using self-attention - MISA, TFN, MELD, MFN, Multimodal InfoMax
• Emotion/Sentiment Recognition in Conversations/Context or Other Dialogue System Works: DialogueRNN, DialogueGCN, BC-LSTM, MIME, CICERO
• Text-to-Audio Generative Models: Tango, TangoFlux
• LLM Evaluation: InstructEval, Reasoning Problems, PuzzleVQA and AlgoPuzzleVQA
• AI Safety/Bias: Red Teaming, Resta, Gender Bias, Cultural Bias
• Retrieval Augmented Generation: CoK, Trust-Score
• Embodied AI: NORA, Emma-X

Updates

🌞 We will be moving to NTU after this summer! 🎓
We're on the lookout for passionate Postdocs and PhD students to join our team. 🚀🧠
Interested? Drop us a message at soujanya [dot] poria [at] gmail [dot] com.
🌟

Keen to Join Our Lab?

Please Read these Slides

If you would like to join our lab, we encourage you to get familiar with the research topics we are interested in. Here are the slides for this.

Intrigued by our logo? Read the story behind it.”