A5043407770- Robotics and Sensor-Based Localization
- Reinforcement Learning in Robotics
- Tactile and Sensory Interactions
- Robot Manipulation and Learning
- Advanced Vision and Imaging
- Human Pose and Action Recognition
- Textile materials and evaluations
- Adversarial Robustness in Machine Learning
- Anomaly Detection Techniques and Applications
- E-commerce and Technology Innovations
- Video Surveillance and Tracking Methods
- Explainable Artificial Intelligence (XAI)
- Consumer Perception and Purchasing Behavior
Stanford University
2018-2019
Contact-rich manipulation tasks in unstructured environments often require both haptic and visual feedback. However, it is non-trivial to manually design a robot controller that combines modalities with very different characteristics. While deep reinforcement learning has shown success control policies for high-dimensional inputs, these algorithms are generally intractable deploy on real robots due sample complexity. We use self-supervision learn compact multimodal representation of our...
Given two consecutive RGB-D images, we propose a model that estimates dense 3D motion field, also known as scene flow. We take advantage of the fact in robot manipulation scenarios, scenes often consist set rigidly moving objects. Our jointly (i) segmentation into an unknown but finite number objects, (ii) trajectories these objects and (iii) object employ hourglass, deep neural network architecture. In encoding stage, RGB depth images undergo spatial compression correlation. decoding...
This paper presents, CoViS, a low-cost real-time amphibious optical computer vision system based on modified 4-connectivity approach that scans acquired frames from video stream for target hue range and then applies few advanced techniques of image processing to the as well. The relatively fast inexpensive provides Cartesian coordinates tracked object, making an option unmanned vehicles acquire, track recognize objects.