Currently working at Qualcomm (San Diego) as a senior engineer for camera system team. I completed my PhD in computer science with focus on computer vision and deep learning from the Center for Research in Computer Vision (CRCV) lab, University of Central Florida (UCF), where I worked with Dr. Yogesh Singh Rawat. My research interests include fundamentals of efficient deep learning applications, including large scale training, image/video generation and segmentation, and sparse annotation training for large datasets. I have previously worked on temporally consistent video segmentation, low cost annotation method, real-time video understanding, driving assistance systems, and robotic manipulation using vision, bringing concepts to practical use.
Prior to starting PhD, I was working at Vision and Graphics Lab at Asian Institute of Technology, Thailand. My primary work was on visual SLAM, tracking, and recognition of vehicles.

NEWS

• We have 1 accepted paper at AAAI 2024!.
• We have 1 accepted paper at CVPR 2023! Project details here.
• I will join Qualcomm as Senior Engineer for camera system team this summer (2023)
• We have 1 accepted paper at NeurIPS 2022! Check project details here.
• Our patent for real-time spatio-temporal activity detection from untrimmed video (US Patent 11468676) has been granted!
• We are organizing Robustness in Sequential Data (ROSE) workshop in CVPR 2022.
• We are hosting the Tiny Actions Challenge as part of ActivityNet workshop at CVPR 2022.
• I will be doing an internship at Qualcomm this summer (2022)
• We have 1 accepted paper at BMVC 2021! for human action generation. Check project page here.
• We have 1 accepted paper at WACV 2021!
• I will be doing an internship at SRI International this summer (2020)

Publications

• Semi-supervised Active Learning for Video Action Detection
  Ayush Singh, Aayush Jung Rana, Akash Kumar, Shruti Vyas, Yogesh Singh Rawat
  AAAI Conference on Artificial Intelligence (AAAI) 2024
  Paper | Github
• Hybrid Active Learning via Deep Clustering for Video Action Detection
  Aayush Jung Rana, Yogesh Singh Rawat
  IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023
  Paper | Github | Resources | Project
• Are all Frames Equal? Active Sparse Learning for Video Action Detection
  Aayush Jung Rana, Yogesh Singh Rawat
  Advances in Neural Information Processing Systems (NeurIPS) 2022
  Paper | Github | Resources | Project
• LARNet: Latent Action Representation for Human Action Synthesis
  Naman Biyani, Aayush Jung Rana, Shruti Vyas, Yogesh Singh Rawat
  British Machine Vision Conference (BMVC) 2021
  Paper | Github | Presentation | Project
• We don’t Need Thousand Proposals: Single Shot Actor-Action Detection in Videos
  Aayush Jung Rana, Yogesh Singh Rawat
  IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2021
  Paper | Github | Presentation | Project
• Gabriella: An Online System for Real-Time Activity Detection in Untrimmed Security Videos
  Mamshad Rizve, Ugur Demir, Praveen Tirupattur, Aayush Jung Rana, Kevin Duarte, Ishan Dave, Yogesh Singh Rawat, and Mubarak Shah
  IEEE International Conference on Pattern Recognition (ICPR) 2020 (Best Paper Award)
  Paper | Presentation | Project
• SSA2D: Single Shot Actor-Action Detection in Videos
  Aayush Jung Rana, Yogesh Singh Rawat
  Student Abstract at the Thirty-Fifth AAAI Conference, 2021
  Paper

Patents

• Methods of Real-Time Spatio-Temporal Activity Detection and Categorization from Untrimmed Video Segments. United States Patent No. 11468676 B2.
  Patent
• Active Sparse Labeling System and Method. United States Patent Application No. 63/514,482 (Filed)