Deep learning using Tensorflow Lite on Raspberry Pi

Name: Deep learning using Tensorflow Lite on Raspberry Pi
Rating: 4.4 (22 reviews)

Power up your Embedded projects with Artificial Intelligence in Python using TF Lite

Created byMuhammad Luqman, Zaheer Ahmed

Last updated 7/2024

English

What you'll learn

Build your own AI Projects
Raspberry Pi 4 based Robot for Computer Vision
Neural Network to classify your Voice
Custom Convolution Network Creation

Course content

3 sections • 63 lectures • 6h 33m total length

How Nerual Networks Work9:28
Non-Linear Function Approximation Understanding3:04
Train a simple neural network with a hidden layer and sigmoid output to approximate the sine function, using mean squared error and the Adam optimizer on 0 to 2 pi.
Trigonometric Function Data Generation9:29
Data Splitting and Normalizing9:08
Deep Learning Model Creation5:35
Model Performance testing and Loss Understanding8:10
Explore model performance testing by generating predictions from a sine model, visualize inputs and predictions, and analyze training and validation losses, mean squared error and mean absolute error across epochs.
Mean Squared Error Graph Understanding5:17
Designing New Improved Model5:29
Model Performance comparisons and Saving5:21
Compare model performance across epochs, noting training loss and error metrics, then save, load, and run predictions on a Keras model for deployment on hardware, with GitHub tracking.
Github Push after Section Completion5:22
Push your development work to a GitHub repository from VS Code, creating a development branch, configuring user details, and pushing notebooks like sign function approximation for project tracking.
Github repository and Resources0:02

Join our free community0:07
Raspberry PI OS setup8:12
Construction of Hardware2:17
Assemble and configure hardware to deploy deep learning models with TensorFlow Lite on a Raspberry Pi, enabling efficient edge inference.
Data Strategy for this project4:01
Producing Custom Data7:07
Raspberry PI SSH Setup using Vscode7:57
Video Saving Script7:46
Create a Raspberry Pi video saving script that captures camera feed, saves video with OpenCV's video writer, and integrates git workflow for data used in deep learning with TensorFlow Lite.
Data Videos Obtaining6:47
Learn to obtain data videos from a Raspberry Pi camera, transfer them to your system using scp, and run a data extractor that saves thousands of frames for processing.
Understanding Frame Extraction Process5:10
Hough Circles Understanding9:17
Number Extraction from Circles6:41
Data Obtaining and Pre Processing9:23
Visualizing Data on Google Colab4:46
3D printing parts Source0:12
Model Architecture3:00
Model Implementation and training9:33
Develop and train a grayscale 128 by 128 convnet with multiple convolution and pooling layers, culminating in a 13-class softmax, using categorical cross entropy and early stopping.
Model Saving2:03
Testing Model6:09
Model Performance Matrices understanding7:32
Post Quantization5:14
TFLite Conversion6:22
Convert your visual calculator model to a tf lite version using the tflite converter, save the model, compare sizes in MB, and run lite interpreter inference for deployment readiness.
TF Lite Model Testing7:24
Real Time Prediction Script8:47
Model Inferencing on recorded Data7:28
Defining Region of Interest7:20
Define a square region of interest in the real-time video feed and crop to this ROI for on-device predictions, moving from recorded video to real-time inference.
Testing ROI Improvements7:31
Explore region-of-interest based circle detection inside a square frame, using targeted crops and lighting-aware predictions, with real-time display logic and Raspberry Pi deployment for machine learning inference.
Raspberry PI Model Inferencing Setup5:21
RPI Inferencing Testing8:23
Equation Building8:28
Number Detection Isolation9:03
Equation Computation13:42
Github Push9:56
Learn to push code frequently on GitHub, implement version control, and document projects with a clear readme. Explore branching and commenting to support Raspberry Pi TensorFlow Lite work.

Understanding Wave Files2:30
Audio Recording Script7:07
Audio Conversion from Float to Integer4:59
Data Recording in batches3:02
Record and batch voice data efficiently for a TensorFlow Lite model on Raspberry Pi by converting floats to integers and preparing labeled batches for training on Google Colab.
Wave file to Binary Tensors5:42
Data Visualizations4:16
Spectrogram Conversion7:22
Data Pre-processing Pipelines6:43
Create labeled spectrogram data by converting raw audio parts to spectrograms, map waveforms to labeled tensors, and prepare training, validation, and testing datasets for a deep learning model.
Model definition and Dataset Splitting6:00
Model Architecture1:50
Discussion Model Parameters and Training Model5:56
Reviewing Training Results8:35
Compare three models by loss and accuracy, address data issues, and select model three with improved validation performance, then run predictions and prepare confusion matrices on test data.
Model Performance Matrix7:02
Tensorflow Lite Conversion and prediction9:13
Input Audio stream pipeline building7:34
Build an input audio stream pipeline on the raspberry pi by reading microphone data, converting to 16-bit, reshaping, and logging predictions from a tensorflow lite model.
TfLite model predictions9:09
LED connections and Blink4:52
Demonstrate LED blinking on a Raspberry Pi using the RPi.GPIO library, wiring LEDs to BCM pins 20 and 21, and run a loop while preparing to add predicting code.
Raspberry Pi Model Predictions6:01
Set up a Raspberry Pi with a microphone to run model predictions and blink LEDs driven by voice labels.
Real Time Audio LED Controlling2:39
Github Push3:22

Requirements

Basic Electronics Understanding
Basic Python Programming
Hardware : Raspberry pi 4
Hardware : 12V Power Bank
Hardware : Raspberry PI Camera V2
Hardware : 2 LEDs ( Red and Green )
Hardware : Bread Board
Hardware : RPI 4 Fan
Hardware : 3D printed Parts
Hardware : Jumper Wires

Description

Course Workflow:

This course is focused on Embedded Deep learning in Python . Raspberry PI 4 is utilized as a main hardware and we will be building practical projects with custom data .

We will start with trigonometric functions approximation . In which we will generate random data and produce a model for Sin function approximation

Next is a calculator that takes images as input and builds up an equation and produces a result .This Computer vision based project is going to be using convolution network architecture for Categorical classification

Another amazing project is focused on convolution network but the data is custom voice recordings . We will involve a little bit of electronics to show the output by controlling our multiple LEDs using own voice .

Unique learning point in this course is Post Quantization applied on Tensor flow models trained on Google Colab . Reducing size of models to 3 times and increasing inferencing speed up to 0.03 sec per input .

Sections :

Non-Linear Function Approximation
Visual Calculator
Custom Voice Controlled Led

Outcomes After this Course : You can create

Deep Learning Projects on Embedded Hardware
Convert your models into Tensorflow Lite models
Speed up Inferencing on embedded devices
Post Quantization
Custom Data for Ai Projects
Hardware Optimized Neural Networks
Computer Vision projects with OPENCV
Deep Neural Networks with fast inferencing Speed

Hardware Requirements

Raspberry PI 4
12V Power Bank
2 LEDs ( Red and Green )
Jumper Wires
Bread Board
Raspberry PI Camera V2
RPI 4 Fan
3D printed Parts

Software Requirements

Python3
Motivated mind for a huge programming Project
----------------------------------------------------------------------------------
Before buying take a look into this course GitHub repository

Who this course is for:

Developers
Electrical Engineers
Artificial Intelligence Enthusiasts

Deep learning using Tensorflow Lite on Raspberry Pi

What you'll learn

Explore related topics

Course content

Non Linear Trigonometric Functions Approximation11 lectures • 1hr 6min

Visual Calculator32 lectures • 3hr 33min

Voice Controlled LEDs20 lectures • 1hr 54min

Requirements

Description

Who this course is for: