Timeframe:
Summer 2019 – Fall 2020
Students:
Purvesh Sharma
Faculty in Collaboration:
Dr. Heping (Fred) Chen
Overview:
The goal is to develop a Deep Learning – Computer Vision approach to detect 3D-objects in RGB-D (Red-Green-Blue-Depth) under different illumination and challenging conditions such as cluttered, impeded and obstructed partial images
Stages
Phase 1:
Deep Convolutional Neural Network (DCNN) is proposed with the usage of RGB-D images, which includes one 2D-image in RGB, and the other in-depth image form. Two methods are proposed, one will be using two parallel DCNN model and another method will be using three parallel DCNN model. Afterward, the parallel models need to be concatenated to get 3D-object detection. Public-available 3D-dataset will be used to evaluate the model, and the results of both of the models are compared
Phase 2:
The architecture uses RGB and Depth images as an input and gives a single output. Most of the researchers used VGG16/19, ResNet, and MobileNet for detection purposes. The architecture is designed to perform a specific task of grasping. For classification using RGB-D architecture, it achieved an average accuracy of 95% with the learning rate of 0.0001 and outperforms the other architectures’ accuracy for limited objects.
Publications:
- P. Sharma, D. Valles, “Deep Convolutional Neural Network Design Approach for 3D Object Detection for Robotic Grasping,” 2020 10th Annual Computing and Communication Workshop and Conference (CCWC), Las Vegas, Nevada, USA, 2020, pp. 0311-0316, doi:10.1109/CCWC47524.2020.9031186
- P. Sharma and D. Valles, “Backbone Neural Network Design of Single Shot Detector from RGB-D Images for Object Detection,” 2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON), New York City, NY, 2020, pp. 0112-0117, doi: 10.1109/UEMCON51285.2020.9298175.
GitHub Channel:
Thesis:
- Sharma, P. (2020). Deep convolutional neural network design approach for 3D-object detection for robotic grasping (Unpublished thesis). Texas State University, San Marcos, Texas.
