Student Talks
Context and Subcategories for Sliding Window Object Recognition
Event Location: NSH 3305Abstract: Object recognition is one of the fundamental challenges in computer vision, where the goal is to identify and localize the extent of object instances within an image. The current de facto standard for building high-performance object category detectors is the sliding window approach. This approach involves scanning an image with a [...]
Lifelong Robotic Object Recognition
Event Location: NSH 3305Abstract: In this thesis, we study the topic of Lifelong Robotic Object Perception. We propose, as a long-term goal, a framework to recognize known objects and to discover unknown objects in the environment as the robot operates, for as long as the robot operates. We build the foundations for Lifelong Robotic Object [...]
Data-Driven Geometric Scene Understanding
Event Location: NSH 3305Abstract: In this thesis proposal, we describe a data-driven approach to leverage repositories of 3D models for scene understanding. Our ability to relate what we see in an image to a large collection of 3D models allows us to transfer information from these models, creating a rich understanding of the scene. We [...]
Robust Natural Language Direction Following through Unknown Environments
Event Location: GHC 4405Abstract: Commanding robots through unconstrained natural language directions is intuitive, flexible, and does not require specialized interfaces or training. Providing this capability would enable effortless coordination in human robot teams that operate in non-specialized environments. However, natural language direction following through unknown environments requires understanding the structure of language, mapping verbs and [...]
Model Recommendation for Action Recognition and Other Applications
Event Location: GHC 4405Abstract: The typical approach to learning based vision has been that for each individual application, classifiers or detectors are learned anew from annotated training data for each specific task. However, the classifiers trained in this way tend to be brittle and highly specialized to the datasets from which they are derived, making [...]
Formulation and Calibration of Fast, Accurate Vehicle Motion Models
Event Location: NSH 3002Abstract: High performance wheeled mobile robots (WMRs) require fast, accurate motion models for planning and control. This thesis addresses two challenges to producing these models: the tradeoff between fidelity and speed in model formulation, and the need for laborious calibration procedures. To address the first challenge, I propose the formulation of “enhanced” [...]
Coherent Scene Understanding with 3D Geometric Reasoning
Event Location: NSH 1305Abstract: Going beyond object detection and semantic segmentation, coherent scene understanding simultaneously considers multiple potential objects and surfaces in the image and reasons over them in a 3D geometric context to derive a coherent interpretation of the scene behind the image, during which many visual ambiguities can be resolved. To achieve this [...]
Multi-agent Pickup and Delivery Planning with Transfers
Event Location: NSH 1305Abstract: Many logistics problems require mobile agents to retrieve and deliver a set of items. For example, postal services and couriers retrieve and deliver letters and packages, while taxis pick up and drop off passengers. These problems are instances of the Pickup and Delivery Problem (PDP), in which a set of mobile [...]
Object Instance Discovery from Scenes of Daily Living
Event Location: NSH 1507Abstract: This thesis tackles the problem of automatically discovering objects from a collection of images from the Activities of Daily Living (ADL) environment. We contribute, 1) a framework for discovering object instances under severe clutter, occlusion, changes of view point, heterogeneity of object appearance and imperfect segmentation; 2) a data-driven approach for [...]
Self-Supervised Learning on Mobile Robots Using Acoustics, Vibration, and Visual Models to Build Rich Semantic Terrain Maps
Event Location: NSH 1305Abstract: Humans and robots would benefit from having rich semantic maps of the terrain in which they operate. Mobile robots equipped with sensors and perception software could build such maps as they navigate through a new environment. This information could then be used by humans or robots for better localization and path [...]