Skip to content

Latest commit

 

History

History
11 lines (8 loc) · 1.37 KB

README.md

File metadata and controls

11 lines (8 loc) · 1.37 KB

grip-net

Real-time frictional Safety Margin Estimation with imaging RGB soft tactile fingertips. Implements a basic fully connected DNN with SOTA performance compared to CNN on training data, and which can outperform SOTA CNN in generalization. The DNN architecture and hyperparameters have not been fine-tuned. The input to the DNN are extracted dot cordinates and radii from preprocessed images. The preprocessing was done by Jingwen Tang, here. The data courtesy of Jingwen Tang, here.

Pre-trained models are available as pytorch lightning checkpoints, and pytorch state dictionaries. Normalized data is available as .csv.

Description

grip-net-scheme Nice figures are from R. Scharff et al. 2022, "Rapid manufacturing of color-based hemispherical soft tactile fingertips." DOI:10.1109/RoboSoft54090.2022.9762136

Performance:

image