Select a calendar:
Filter December Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Events for December 03, 2021
-
PhD Thesis Proposal - Arka Sadhu
Fri, Dec 03, 2021 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
University Calendar
Ph.D. Thesis Proposal - Arka Sadhu
Friday, Dec 3rd, 2021: 12pm-2pm
Title: Grounding Language in Images and Videos
Thesis Committee members: Prof. Ram Nevatia, Prof. Xiang Ren, Prof. Yan Liu, Prof. Stefanos Nikolaidis, Prof. Toby Mintz.
Abstract: Language grounding in images and videos -- the task of associating linguistic symbols to perceptual experiences and actions -- is fundamental to developing multi-modal models which can understand and jointly reason over images, videos and text.
It has garnered wide interest from multiple disciplines such as computer vision, natural language processing, and robotics. An essential element in this space involves formulating tasks that investigate a particular phenomenon inherent in image or video understanding in isolation, thereby encouraging the community to develop more robust models. In this thesis proposal, I will articulate four vision-language tasks developed during the course of my Ph.D., namely, grounding unseen words, spatio-temporal localization of entities in a video, video question-answering, and visual semantic role labeling in videos. For each of these tasks, I will further discuss the development of corresponding datasets, evaluation protocols, and model frameworks.
Zoom Link: https://usc.zoom.us/j/92383912262?pwd=N25ETlRMVFRiWTlKdGxtN09UVHhlQT09WebCast Link: https://usc.zoom.us/j/92383912262?pwd=N25ETlRMVFRiWTlKdGxtN09UVHhlQT09
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
PhD Thesis Proposal - Chen-Yu Wei
Fri, Dec 03, 2021 @ 01:00 PM - 02:30 PM
Thomas Lord Department of Computer Science
University Calendar
Time: 1:00-2:30pm, December 3rd
Committee: Haipeng Luo (host), Rahul Jain, David Kempe, Vatsal Sharan, Jiapeng Zhang
Title: Robust and Adaptive Online Reinforcement Learning
Abstract: Online reinforcement learning (RL) studies how an agent learns to behave in an unknown environment from scratch. In this thesis, I focus on the theoretical foundations of this learning paradigm, with emphasis on designing algorithms that are robust to the non-stationarity of the environment, where the non-stationarity may come from natural drift, adversarial manipulation, or the existence of other agents. While being robust, most of our algorithms are also "adaptive" at the same time in the sense that they do not sacrifice nice performance guarantees if the environment happens to be stationary. More broadly speaking, the performance of our algorithms automatically scale with some intrinsic properties that reflect the difficulty of the problem.
For future work, I plan to characterize the fundamental limit of RL in large state space, a central topic in theoretical RL. We hope to answer the following questions: "what are the minimal assumptions to be made so that RL algorithms can find near-optimal policies with polynomial number of samples", and the similar question under the restriction of "polynomial computational time".WebCast Link: https://usc.zoom.us/j/96695544670?pwd=VnZJUzRLam9scVpHbFRTYUVmQlk4Zz09
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
