Deeplab ADE20K inference (Tensorflow 2)

Python program to visualize the results from the Deeplab model (trained on ADE20K dataset).

Required Python packages

• numpy: For the array calculations.
• tensorflow: Tensorflow 2.
• Pillow: To read the image.
• Matplotlib: To plot the results.

HOW TO USE

• Download the required packages.
• Download the model from the model zoo. ONLY model trained in ADE20K
• Change the name of the path to the model (model_path) and the image (img_path) you want to test.
• Run the program.

Example

For an example on how to visualize the inference of DeepLab model (ADE20K) run inferenceDeeplabADE20K.py script.

git clone https://github.com/ibaiGorordo/Deeplab-ADE20K-Inference.git
python inferenceDeeplabADE20K.py