Saumya Gupta

selected publications

1. 

   SmartEdit: Editing-driven Engagement Prediction and Enhancement of Short-Videos

   Saumya Gupta, Ishita Dasgupta, Stefano Petrangeli, and 1 more author

   In 2025 IEEE International Conference on Multimedia and Expo (ICME) 2025

   Bib PDF

   @inproceedings{11209033,
     author = {Gupta, Saumya and Dasgupta, Ishita and Petrangeli, Stefano and Sarkhel, Somdeb},
     booktitle = {2025 IEEE International Conference on Multimedia and Expo (ICME)},
     title = {SmartEdit: Editing-driven Engagement Prediction and Enhancement of Short-Videos},
     year = {2025},
     pages = {1-11},
     keywords = {Measurement;Systematics;Social networking (online);Large language models;Buildings;Streaming media;Videos;short-video;dataset;engagement prediction;editing improvement},
     doi = {10.1109/ICME59968.2025.11209033}
   }

2. 

   TopoDiffusionNet: A Topology-aware Diffusion Model

   Saumya Gupta, Dimitris Samaras, and Chao Chen

   International Conference on Learning Representations (ICLR) 2025

   arXiv Bib Code Poster Slides

   @article{gupta2024topodiffusionnet,
     title = {TopoDiffusionNet: A Topology-aware Diffusion Model},
     author = {Gupta, Saumya and Samaras, Dimitris and Chen, Chao},
     journal = {International Conference on Learning Representations (ICLR)},
     year = {2025},
   }

3. 

   Topology-aware Uncertainty for Image Segmentation

   Saumya Gupta, Yikai Zhang, Xiaoling Hu, and 2 more authors

   Advances in Neural Information Processing Systems (NeurIPS) 2023

   arXiv Bib HTML Code Poster Slides

   @article{gupta2024topology,
     title = {Topology-aware Uncertainty for Image Segmentation},
     author = {Gupta, Saumya and Zhang, Yikai and Hu, Xiaoling and Prasanna, Prateek and Chen, Chao},
     journal = {Advances in Neural Information Processing Systems (NeurIPS)},
     volume = {36},
     year = {2023},
   }

4. 

   Learning Topological Interactions for Multi-Class Medical Image Segmentation

   Saumya Gupta, Xiaoling Hu, James Kaan, and 8 more authors

   In European Conference on Computer Vision (ECCV) 2022

   arXiv Bib HTML Code Poster Slides

   @inproceedings{gupta2022learning,
     title = {Learning Topological Interactions for Multi-Class Medical Image Segmentation},
     author = {Gupta, Saumya and Hu, Xiaoling and Kaan, James and Jin, Michael and Mpoy, Mutshipay and Chung, Katherine and Singh, Gagandeep and Saltz, Mary and Kurc, Tahsin and Saltz, Joel and others},
     booktitle = {European Conference on Computer Vision (ECCV)},
     pages = {701--718},
     year = {2022},
     organization = {Springer},
   }