|
Ashish Goswami I'm a third year PhD student (TCS Research Fellow) at Yardi School of AI@IIT-Delhi, where I'm working on object-centric generation and editing of images using Diffusion Models, advised by Prof. Parag Singla I completed my Bachelors in Electronics and Communication Engineering from IIIT-Guwahati, during which I did a few internships centered around Large Scale ML and generative models at Delhivery and Spyne.ai |
|
ResearchI'm broadly interested in building reasoning capabilities in multi-modal generative models like Text-to-image diffusion and Text driven image-editing models. Recent work on object centric representation learning has some pretty neat ideas which I'd like explore. |
|
Image Manipulation via Multi-Hop Instructions -- A New Dataset and Weakly-Supervised Neuro-Symbolic Approach
Harman Singh, Poorva Garg, Mohit Gupta, Kevin Shah, Ashish Goswami, Satyam Modi, Arnab Kumar Mondal, Dinesh Khandelwal, Parag Singla, Dinesh Garg, EMNLP, 2023 video / arXiv New datasets and a modular method for weakly-supervised instruction guided image manipulations. |
|
GraPE: A Generate-Plan-Edit Framework for Compositional T2I Synthesis
Ashish Goswami, Satyam Modi, Satyam Rishi Deshineni, Harman Singh, Prathosh A. P, Parag Singla, Arxiv Preprint, 2024 Project Page / arXiv / Code An unifying and generic framework for improving semantic alignemt of T2I models by post-hoc alignment of generated images via iterating editing. |
TA Duties |
|
| Sem I - 2022-23 | COL671: Intro to Principles of AI |
| Sem II - 2022-23 | AIL861: Special topics in Applications |
| Sem I - 2023-24 | COL774: Machine Learning |
| Sem II - 2023-24 | COL775: Deep Learning |
| Sem I - 2024-25 | COL828: Advance Computer Vision |