|
--- |
|
license: mit |
|
datasets: |
|
- biplab2008/viscositydset |
|
language: |
|
- en |
|
metrics: |
|
- name: r_squared |
|
type: r_squared |
|
value: 0.906 |
|
tags: |
|
- chemistry |
|
- pytorch |
|
- video-regression |
|
pipeline_tag: image-feature-extraction |
|
--- |
|
|
|
# Viscosity Regression Model |
|
Regress the viscosity value [cP] directly from videos using CNN3D and Feedforward NN; no need for hand-crafted feature engineering. |
|
|
|
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6632618508a36fda54f27609/w_HLzfUAq33Qo4gS2MsXw.png) |
|
|
|
|
|
# Example of Video Frames |
|
<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/6632618508a36fda54f27609/o9oitVnhhZpAG7OrT0LuE.mp4"></video> |
|
|
|
# Model Architecture |
|
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6632618508a36fda54f27609/PHnchHN6_dAC9o8bo9Tvs.png) |
|
|
|
## Train & Test R2 |
|
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6632618508a36fda54f27609/7q6Xrx8upSIwIwgC1wEEw.png) |
|
|