Computer Vision

Sequential skeleton and average foot pressure data for normal and five pathological gaits (i.e., antalgic, lurching, steppage, stiff-legged, and Trendelenburg) were simultaneously collected. The skeleton data were collected by using Azure Kinect (Microsoft Corp. Redmond, WA, USA). The average foot pressure data were collected by GW1100 (GHIWell, Korea). 12 healthy subjects participated in data collection. They simulated the pathological gaits under strict supervision. A total of 1,440 data instances (12 people x 6 gait types x 20 walkings) were collected.

We propose a novel high-resolution dataset named, “Dataset for Indian Road Scenarios (DIRS21)” for developing perception systems for advanced driver assistance systems.

Opportunity++ is a precisely annotated dataset designed to support AI and machine learning research focused on the multimodal perception and learning of human activities (e.g. short actions, gestures, modes of locomotion, higher-level behavior).

We design a solution to achieve coordinated localization between two unmanned aerial vehicles (UAVs) using radio and camera perception. We achieve the localization between the UAVs in the context of solving the problem of UAV Global Positioning System (GPS) failure or its unavailability. Our approach allows one UAV with a functional GPS unit to coordinate the localization of another UAV with a compromised or missing GPS system. Our solution for localization uses a sensor fusion and coordinated wireless communication approach.

To study the driver's behavior in real traffic situations, we conducted experiments using an instrumented vehicle, which comprises:

(i) a camera, installed above the vehicle's side window and oriented toward the driver, and (ii) a Mobile Digital Video Recorder (MDVR).

Surveillance video captured by Multi-intensity infrared illuminator.

GT(ground-truths) :bounding boxes of 'person' in channel 2,4 and 6 by following the Pascal VOC format.

The Dasha River dataset was collected by a USV sailing along the Dasha River in Shenzhen, China. Visual images in the dataset were extracted from two videos taken from a USV perspective, with a resolution of 1920×1080 pixels. Totally 360 images were obtained after screening, and all labels were manually annotated.