AndreasLH commited on
Commit
e3371e6
·
1 Parent(s): 86cda90
Files changed (2) hide show
  1. Dockerfile +1 -1
  2. app.py +3 -3
Dockerfile CHANGED
@@ -1,5 +1,5 @@
1
  # Base image, have to use the full version to use the git features
2
- FROM python:3.11
3
  # https://huggingface.co/docs/hub/spaces-sdks-docker-first-demo
4
 
5
  # RUN apt-get install -y git
 
1
  # Base image, have to use the full version to use the git features
2
+ FROM python:3.10
3
  # https://huggingface.co/docs/hub/spaces-sdks-docker-first-demo
4
 
5
  # RUN apt-get install -y git
app.py CHANGED
@@ -133,7 +133,7 @@ if __name__ == "__main__":
133
  return model
134
 
135
  title = 'Weak Cube R-CNN'
136
- description = "This showcases the different our model [[`Weak Cube RCNN`](https://arxiv.org/abs/2504.13297). To create Weak Cube RCNN, we modify the framework by replacing its 3D loss functions with ones based solely on 2D annotations. Our methods rely heavily on external, strong generalised deep learning models to infer spatial information in scenes. Experimental results show that all models perform comparably to an annotation time-equalised Cube R-CNN, whereof the pseudo ground truth method achieves the highest accuracy. The results show the methods' ability to understand scenes in 3D, providing satisfactory visual results. Although not precise enough for centimetre accurate measurements, the method provide a solid foundation for further research. \n Check out the code on [GitHub](https://github.com/AndreasLH/Weak-Cube-R-CNN)"
137
 
138
 
139
  demo = gr.Interface(
@@ -151,5 +151,5 @@ if __name__ == "__main__":
151
  )
152
 
153
 
154
- # demo.launch(server_name="0.0.0.0", server_port=7860)
155
- demo.launch()
 
133
  return model
134
 
135
  title = 'Weak Cube R-CNN'
136
+ description = "This showcases the different our model [`Weak Cube RCNN`](https://arxiv.org/abs/2504.13297). To create Weak Cube RCNN, we modify the framework by replacing its 3D loss functions with ones based solely on 2D annotations. Our methods rely heavily on external, strong generalised deep learning models to infer spatial information in scenes. Experimental results show that all models perform comparably to an annotation time-equalised Cube R-CNN, whereof the pseudo ground truth method achieves the highest accuracy. The results show the methods' ability to understand scenes in 3D, providing satisfactory visual results. Although not precise enough for centimetre accurate measurements, the method provide a solid foundation for further research. \n Check out the code on [GitHub](https://github.com/AndreasLH/Weak-Cube-R-CNN)"
137
 
138
 
139
  demo = gr.Interface(
 
151
  )
152
 
153
 
154
+ demo.launch(server_name="0.0.0.0", server_port=7860)
155
+ # demo.launch()