Post by Tajamul Ashraf

✨ Double Delight: #ICCV2025 & #MICCAI2025! 🎉 The past year was truly special for me. I graduated from my master’s and had the incredible opportunity to work at Microsoft Research. From solving fundamental problems to discovering that coffee counts as a food group, every day was a mix of learning, laughter, and legendary lunch breaks. 🩺 MICCAI 2025 📄 Phase-Informed Tool Segmentation for Manual Small-Incision Cataract Surgery In this work, we introduce the first large-scale dataset for MSICS, a low-cost and high-impact cataract surgery technique, and propose ToolSeg, a phase-aware, semi-supervised framework for robust tool segmentation. Link: https://lnkd.in/d8u6EMbv Joint work with Bhuvan Sachdeva, Naren Akash R J, Simon Mueller, Thomas Schultz, Maximilian W. M. Wintergerst, Niharika Singri Prasad, Kaushik Murali, and Mohit Jain at Microsoft Research, Sankara Eye Hospital, and The University of Bonn. 📷 ICCV 2025 📄 TITAN: Query-Token based Domain Adaptive Adversarial Learning TITAN addresses source-free domain adaptive object detection using a novel query-token adversarial strategy and variance-based domain partitioning. Link: https://lnkd.in/d7474KjG This work was in collaboration with Janibul Bashir from Gaash Research Lab, National Institute of Technology Srinagar, incredibly proud to see our joint effort make it to ICCV! Both of these projects shaped me deeply, from vision theory to real-world clinical relevance. I am immensely thankful to my mentors, co-authors, and the research communities that continue to inspire me. Here’s to more meaningful research ahead! 🌱 #arXiv #ICCV2025 #MICCAI2025 #ComputerVision #MedicalImaging #DomainAdaptation #ToolSeg #TITAN #AIforHealthcare #MicrosoftResearch #GaashLab #MSICS #DeepLearning #AcademicLife #ResearchMilestone