I am a senior research engineer at Qualcomm Research in Amsterdam, working on computer vision problems using machine learning. Before joining Qualcomm, I was a postdoctoral researcher at the Intelligent Systems Lab Amsterdam of the University of Amsterdam, supervised by Cees Snoek. During postdoc, I mainly worked on action recognition in videos. During my PhD, I worked with Hervé Jegou, Patrick Bouthemy and Patrick Gros. I received my PhD in April'2014. Objective of my thesis was to improve image and video representation for visual recognition.
- Embeddings available. For "Objects2action: Classifying and localizing actions without any video example", object representations and embeddings are available on the project page
- Features available. For "What do 15,000 object categories tell us about classifying and localizing actions?", features and other details are available on the project page
- Code available. For "APT: Action localization Proposals from dense Trajectories" (.pdf) python code and pre-computed tubes available on github
- Accepted paper. ICCV 2015, Objects2action: Classifying and localizing actions without any video example
- Accepted paper. BMVC 2015, oral presentation, APT: Action localization Proposals from dense Trajectories
- Accepted paper. CVPR 2015, What do 15,000 Object Categories Tell Us About Classifying and Localizing Actions?
Teaching at UvA
- MSc course Visual Search Engine, 2015