- ICNS Home
- Projects
- Current PhD Projects
- Automatic Real-Time Recognition of Objects in Photos and Videos on Edge Devices
Automatic Real-Time Recognition of Objects in Photos and Videos on Edge Devices
Research Student
Research Project Details
Neutral networks, especially deep neural networks, is the major class of machine learning techniques that are increasingly being used in real-time object detection. Classification of objects, static region detection, and clip and key-frame detection errors needs to be accounted while detecting real time objects. Much of the current work in this area has been in the domain of object classification. Recently, there is an increased interest in a new approach to detect object using YOLO. Object detection is considered as a regression problem to spatially separate bounding boxes and associated class probabilities. A single neural network predicts bounding boxes and class probabilities directly from full images in one evaluation. Since the whole detection pipeline is considered as a single network, it can be optimized end-to-end directly on detection performance, which is the main focus of this project.
About ICNS
Partnership
Research
Projects
-
- Automatic Real-Time Recognition of Objects in Photos and Videos on Edge Devices
- Cortical Microcircuit Models of Predictive Processing for Neuromorphic Processors
- Neuromorphic Perception for Greenhouse Technology
- Hardware Implementation of Unsupervised Feature Extraction Algorithms
- Brain-Inspired Neuromorphic Compute Architectures for Edge Computing
- Odor Identification and Odor Source Localization using Event Based Gas Sensing
Resources
Master's Program
Opportunities
Events
News
Contact