Robotic Inference

You've made it to the most fun part of this lesson; project submission! We eagerly look forward to your work and what you have taken away from this development exercise.

Here are the project steps:

1) Using the supplied data in the digits image, create a network that achieves at least 75 percent accuracy and an inference time of less than 10 ms. This can be done by running the evaluate command. Be sure to take a screenshot of the results and add the date and your name to the image.

2) Determine what you want your robotic inference idea to be. Collect your own data that fits with what your model is trying to accomplish. Train a network that produces meaningful results.

3) Document the project following the provided structure. Be sure to include and talk about both networks you worked on. You could use this template

When you have completed the steps above and have cross checked them to the rubric here, go ahead and submit your work!

Be sure to include the following items in your project submission:

• A write up in PDF format that addresses the rubric points. Be sure to include sample images of the items you are classifying, charts, graphs, etc.
• A photo of your terminal output after running the evaluate command with your name and the date watermarked on it. This can also just be included in your write up.
• The model for the supplied data set. This includes: deploy.prototxt, labels.txt, mean.binaryproto, your_model.caffemodel, solver.prototxt and train_val.prototxt.

Feedback

Please fill out the Project Feedback Form after you have completed the project.