You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This Learning Path is about TinyML. It serves as a starting point for learning how cutting-edge AI technologies may be used on even the smallest devices, making Edge AI more accessible and efficient. You will learn how to set up your host machine and target device to facilitate compilation and ensure smooth integration across devices.
9
+
## TinyML
10
10
11
-
In this section, you get an overview of the domain with real-life use-cases and available devices.
12
11
13
-
## Overview
14
-
TinyML represents a significant shift in machine learning deployment. Unlike traditional machine learning, which typically depends on cloud-based servers or high-performance hardware, TinyML is tailored to function on devices with limited resources, constrained memory, low power, and less processing capabilities. TinyML has gained popularity because it enables AI applications to operate in real-time, directly on the device, with minimal latency, enhanced privacy, and the ability to work offline. This shift opens up new possibilities for creating smarter and more efficient embedded systems.
12
+
This Learning Path is about TinyML. It is a starting point for learning how innovative AI technologies can be used on even the smallest of devices, making Edge AI more accessible and efficient. You will learn how to set up your host machine and target device to facilitate compilation and ensure smooth integration across devices.
13
+
14
+
This section provides an overview of the domain with real-life use cases and available devices.
15
+
16
+
TinyML represents a significant shift in Machine Learning deployment. Unlike traditional Machine Learning, which typically depends on cloud-based servers or high-performance hardware, TinyML is tailored to function on devices with limited resources, constrained memory, low power, and fewer processing capabilities.
17
+
18
+
TinyML has gained popularity because it enables AI applications to operate in real-time, directly on the device, with minimal latency, enhanced privacy, and the ability to work offline. This shift opens up new possibilities for creating smarter and more efficient embedded systems.
15
19
16
20
### Benefits and applications
17
21
18
-
The advantages of TinyML match up well with the Arm architecture, which is widely used in IoT, mobile devices, and edge AI deployments.
22
+
The benefits of TinyML align well with the Arm architecture, which is widely used in IoT, mobile devices, and edge AI deployments.
19
23
20
-
Here are some key benefits of TinyML on Arm:
24
+
Here are some of the key benefits of TinyML on Arm:
21
25
22
26
23
27
-**Power Efficiency**: TinyML models are designed to be extremely power-efficient, making them ideal for battery-operated devices like sensors, wearables, and drones.
24
28
25
-
-**Low Latency**: Because the AI processing happens on-device, there's no need to send data to the cloud, reducing latency and enabling real-time decision-making.
29
+
-**Low Latency**: AI processing happens on-device, so there is no need to send data to the cloud, which reduces latency and enables real-time decision-making.
26
30
27
-
-**Data Privacy**: With on-device computation, sensitive data remains local, providing enhanced privacy and security. This is particularly crucial in healthcare and personal devices.
31
+
-**Data Privacy**: With on-device computation, sensitive data remains local, providing enhanced privacy and security. This is a priority in healthcare and personal devices.
28
32
29
-
-**Cost-Effective**: Arm devices, which are cost-effective and scalable, can now handle sophisticated machine learning tasks, reducing the need for expensive hardware or cloud services.
33
+
-**Cost-Effective**: Arm devices, which are cost-effective and scalable, can now handle sophisticated Machine Learning tasks, reducing the need for expensive hardware or cloud services.
30
34
31
35
-**Scalability**: With billions of Arm devices in the market, TinyML is well-suited for scaling across industries, enabling widespread adoption of AI at the edge.
32
36
33
-
TinyML is being deployed across multiple industries, enhancing everyday experiences and enabling groundbreaking solutions. The table below contains a few examples of TinyML applications.
37
+
TinyML is being deployed across multiple industries, enhancing everyday experiences and enabling groundbreaking solutions. The table below shows some examples of TinyML applications.
34
38
35
-
| Area | Device, Arm IP | Description |
36
-
| ------ | ------- | ------------ |
37
-
| Healthcare | Fitbit Charge 5, Cortex-M | Monitor vital signs such as heart rate, detect arrhythmias, and provide real-time feedback. |
38
-
| Agriculture | OpenAg, Cortex-M | Monitor soil moisture and optimize water usage. |
39
-
| Home automation | Arlo, Cortex-A | Detect objects and people, trigger alerts or actions while saving bandwidth and improving privacy. |
40
-
| Industrial IoT | Siemens, Cortex-A | Analyze vibration patterns in machinery to predict when maintenance is needed and prevent breakdowns. |
41
-
| Wildlife conservation | Conservation X, Cortex-M | Identify animal movements or detect poachers in remote areas without relying on external power sources. |
39
+
| Area | Device, Arm IP | Description|
40
+
| ------ | ------- | ------------|
41
+
| Healthcare | Fitbit Charge 5, Cortex-M | To monitor vital signs such as heart rate, detect arrhythmias, and provide real-time feedback. |
42
+
| Agriculture | OpenAg, Cortex-M | To monitor soil moisture and optimize water usage. |
43
+
| Home automation | Arlo, Cortex-A | To detect objects and people, trigger alerts or actions while saving bandwidth and improving privacy. |
44
+
| Industrial IoT | Siemens, Cortex-A | To analyze vibration patterns in machinery to predict when maintenance is needed and prevent breakdowns. |
45
+
| Wildlife conservation | Conservation X, Cortex-M | To identify animal movements or detect poachers in remote areas without relying on external power sources. |
42
46
43
47
### Examples of Arm-based devices
44
48
45
-
There are many Arm-based devices you can use for TinyML projects. Some of them are listed below, but the list is not exhaustive.
49
+
There are many Arm-based devices that you can use for TinyML projects. Some of these are detailed below, but the list is not exhaustive.
46
50
47
51
#### Raspberry Pi 4 and 5
48
52
49
53
Raspberry Pi single-board computers are excellent for prototyping TinyML projects. They are commonly used for prototyping machine learning projects at the edge, such as in object detection and voice recognition for home automation.
50
54
51
55
#### NXP i.MX RT microcontrollers
52
56
53
-
NXP i.MX RT microcontrollers are low-power microcontrollers that can handle complex TinyML tasks while maintaining energy efficiency, making them ideal for applications like wearable healthcare devices and environmental sensors.
57
+
NXP i.MX RT microcontrollers are low-power microcontrollers that can handle complex TinyML tasks while maintaining energy efficiency. This makes them ideal for applications like wearable healthcare devices and environmental sensors.
54
58
55
59
#### STM32 microcontrollers
56
60
@@ -66,4 +70,4 @@ In addition to hardware, there are software platforms that can help you build Ti
66
70
67
71
Edge Impulse offers a suite of tools for developers to build and deploy TinyML applications on Arm-based devices. It supports devices like Raspberry Pi, Arduino, and STMicroelectronics boards.
68
72
69
-
Now that you have an overview of the subject, move on to the next section where you will set up an environment on your host machine.
73
+
Now that you have an overview of the subject, you can move on to the next section where you will set up an environment on your host machine.
Copy file name to clipboardExpand all lines: content/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/_index.md
+7-7Lines changed: 7 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -7,18 +7,18 @@ cascade:
7
7
8
8
minutes_to_complete: 40
9
9
10
-
who_is_this_for: This is an introductory topic for developers, engineers, and data scientists who are new to TinyML and interested in exploring its potential for edge AI. You will learn how to get started using PyTorch and ExecuTorch for TinyML.
10
+
who_is_this_for: This is an introductory topic for developersand data scientists new to Tiny Machine Learning (TinyML) who want to explore its potential using PyTorch and ExecuTorch.
11
11
12
12
learning_objectives:
13
-
- Identify how TinyML is different from other AI domains.
14
-
- Understand the benefits of deploying AI models on Arm-based edge devices.
15
-
- Select Arm-based devices for TinyML.
16
-
- Install and configure a TinyML development environment using ExecuTorch and the Corstone-320 FVP.
13
+
- Describe what differentiates TinyML from other AI domains.
14
+
- Describe the benefits of deploying AI models on Arm-based edge devices.
15
+
- Identify suitable Arm-based devices for TinyML applications.
16
+
- Set up and configure a TinyML development environment using ExecuTorch and Corstone-320 FVP.
17
17
18
18
prerequisites:
19
-
- Basic knowledge of machine learning concepts.
19
+
- Basic knowledge of Machine Learning concepts.
20
20
- A Linux host machine or VM running Ubuntu 22.04 or higher.
21
-
- A [Grove Vision AI Module](https://wiki.seeedstudio.com/Grove-Vision-AI-Module/) or an Arm license to run the Corstone-300 Fixed Virtual Platform (FVP).
21
+
- A [Grove Vision AI Module](https://wiki.seeedstudio.com/Grove-Vision-AI-Module/) or an Arm license to run the Corstone-320 Fixed Virtual Platform (FVP).
Copy file name to clipboardExpand all lines: content/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/build-model-8.md
+5-5Lines changed: 5 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -10,9 +10,9 @@ layout: "learningpathall"
10
10
11
11
## Define a small neural network using Python
12
12
13
-
With the development environment ready, you can create a simple PyTorch model to test the setup.
13
+
With your development environment set up, you can create a simple PyTorch model to test the setup.
14
14
15
-
This example defines a small feedforward neural network for a classification task. The model consists of 2 linear layers with ReLU activation in between.
15
+
This example defines a small feedforward neural network for a classification task. The model consists of two linear layers with ReLU activation in between.
16
16
17
17
Use a text editor to create a file named `simple_nn.py` with the following code:
From the Arm Examples directory, you build an embedded Arm runner with the `.pte` included. This allows you to get the most performance out of your model, and ensures compatibility with the CPU kernels on the FVP. Finally, generate the executable `arm_executor_runner`.
74
+
From the Arm Examples directory, you can build an embedded Arm runner with the `.pte` included. This allows you to optimize the performance of your model, and ensures compatibility with the CPU kernels on the FVP. Finally, generate the executable `arm_executor_runner`.
Copy file name to clipboardExpand all lines: content/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/env-setup-5.md
+8-8Lines changed: 8 additions & 8 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -8,11 +8,11 @@ weight: 3
8
8
layout: "learningpathall"
9
9
---
10
10
11
-
In this section, you will prepare a development environment to compile a machine learning model. These instructions have been tested on Ubuntu 22.04, 24.04 and on Windows Subsystem for Linux (WSL).
11
+
In this section, you will prepare a development environment to compile a Machine Learning model. These instructions have been tested on Ubuntu 22.04, 24.04, and on Windows Subsystem for Linux (WSL).
12
12
13
13
## Install dependencies
14
14
15
-
Python3 is required and comes installed with Ubuntu, but some additional packages are needed.
15
+
Python3 is required and comes installed with Ubuntu, but some additional packages are needed:
Run the commands below to set up the ExecuTorch internal dependencies.
43
+
Run the commands below to set up the ExecuTorch internal dependencies:
44
44
45
45
```bash
46
46
git submodule sync
@@ -50,7 +50,7 @@ git submodule update --init
50
50
```
51
51
52
52
{{% notice Note %}}
53
-
If you run into an issue of `buck` running in a stale environment, reset it by running the following instructions.
53
+
If you run into an issue of `buck` running in a stale environment, reset it by running the following instructions:
54
54
55
55
```bash
56
56
ps aux | grep buck
@@ -70,8 +70,8 @@ executorch 0.6.0a0+3eea1f1
70
70
71
71
## Next Steps
72
72
73
-
Your next steps depends on the hardware you have.
73
+
Your next steps depend on your hardware.
74
74
75
-
If you have the Grove Vision AI Module proceed to [Set up the Grove Vision AI Module V2](/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/setup-7-grove/).
75
+
If you have the Grove Vision AI Module, proceed to [Set up the Grove Vision AI Module V2 Learning Path](/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/setup-7-grove/).
76
76
77
-
If you don't have the Grove Vision AI Module, you can use the Corstone-300 FVP instead, proceed to [Set up the Corstone-300 FVP](/learning-paths/microcontrollers/introduction-to-tinyml-on-arm/env-setup-6-fvp/).
77
+
If you do not have the Grove Vision AI Module, you can use the Corstone-320 FVP instead. See the Learning Path [Set up the Corstone-320 FVP](/learning-paths/microcontrollers/introduction-to-tinyml-on-arm/env-setup-6-fvp/).
Copy file name to clipboardExpand all lines: content/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/env-setup-6-FVP.md
+4-4Lines changed: 4 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -10,25 +10,25 @@ layout: "learningpathall"
10
10
11
11
## Corstone-320 FVP Setup for ExecuTorch
12
12
13
-
Navigate to the Arm examples directory in the ExecuTorch repository.
13
+
Navigate to the Arm examples directory in the ExecuTorch repository:
14
14
15
15
```bash
16
16
cd$HOME/executorch/examples/arm
17
17
./setup.sh --i-agree-to-the-contained-eula
18
18
```
19
19
20
-
After the script has finished running, it prints a command to run to finalize the installation. This will add the FVP executable's to your path.
20
+
After the script has finished running, it prints a command to run to finalize the installation. This step adds the FVP executables to your system path.
Test that the setup was successful by running the `run.sh` script for Ethos-U85, which is the target device for Corstone-320.
26
+
Test that the setup was successful by running the `run.sh` script for Ethos-U85, which is the target device for Corstone-320:
27
27
28
28
```bash
29
29
./examples/arm/run.sh --target=ethos-u85-256
30
30
```
31
31
32
32
You will see a number of examples run on the FVP.
33
33
34
-
This confirms the installation, and you can proceed to the next section[Build a Simple PyTorch Model](/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/build-model-8/).
34
+
This confirms the installation, so you can now proceed to the Learning Path[Build a Simple PyTorch Model](/learning-paths/embedded-and-microcontrollers/introduction-to-tinyml-on-arm/build-model-8/).
1. Download and extract the latest Edge Impulse firmware
29
29
Grove Vision V2 [Edge impulse Firmware](https://cdn.edgeimpulse.com/firmware/seeed-grove-vision-ai-module-v2.zip).
30
30
31
-
2. Connect the Grove - Vision AI Module V2 to your computer using the USB-C cable.
31
+
2. Connect the Grove Vision AI Module V2 to your computer using the USB-C cable.
32
32
33
33
34
34
@@ -41,5 +41,10 @@ Ensure the board is properly connected and recognized by your computer.
41
41
```console
42
42
./flash_linux.sh
43
43
```
44
+
You have now set up the board successfully. In the next section, you will learn how to use the functionality in the ExecuTorch repository for TinyML, using a hardware emulator.
45
+
46
+
{{% notice Note %}}
47
+
In the next Learning Path in this series, you will incorporate the board into the workflow, running workloads on real hardware.
48
+
{{% /notice %}}
44
49
45
50
Continue to the next page to build a simple PyTorch model.
0 commit comments