Tensorflow Lite for deep learning on the Raspberry Pi

Python-based embedded deep learning is the main topic of this course. The main piece of hardware is a Raspberry PI 4, and we’ll use personal information to make useful apps.

Trigonometric function approximation will be our first step. In this section, we’ll make some random data and a model to get close to the sin function.

The following calculator creates an equation using input from photos and outputs the result. A computer vision-based project will be used in this computer vision-based project for category categorization.

Convolution networks are at the center of yet another incredible piece of research, but this time the data utilized is made-to-order voice recordings. We’ll show the results by speaking commands to a number of LEDs through a few circuits.

The application of Post Quantization to Tensor Flow Models trained on Google Colab is a special learning point in this course. The speed of inference is sped up to 0.03 seconds per input, and the size of the model is cut by three orders of magnitude.

Results Following This Course: You Can

1. Projects for Deep Learning on Embedded Hardware
2. You can create Tensorflow Lite models from your existing models.
3. Embedded device inferencing should be expedited.
4. After Quantification
5. Projects Using Custom Data for AI
6. Adaptive Neural Networks for Hardware
7. OpenCV-based computer vision projects
8. Deep Learning Networks with Quick Inferencing