TensorFlow_Lite_Segmentation_Jetson-Nano

TensorFlow Lite Unet running on a Jetson Nano

A fast C++ implementation of TensorFlow Lite Unet on a Jetson Nano.
Once overclocked to 2015 MHz, the app runs at 11 FPS.
Special made for a Jetson Nano see Q-engineering deep learning examples

---

Papers: https://arxiv.org/abs/1606.00915
Training set: VOC2017
Size: 257x257

---

Benchmark.

CPU 2015 MHz	GPU 2015 MHz	CPU 1479 MHz	GPU 1479 MHZ	RPi 4 64os 1950 MHz
11 FPS	9.1 FPS	9 FPS	8.3 FPS	7.2 FPS

---

Dependencies.

To run the application, you have to:

• TensorFlow Lite framework installed. Install TensorFlow Lite
  
• Optional OpenCV installed. Install OpenCV 4.5
  
• Code::Blocks installed. ($ sudo apt-get install codeblocks)

---

Installing the app.

To extract and run the network in Code::Blocks
$ mkdir MyDir
$ cd MyDir
$ wget https://github.com/Qengineering/TensorFlow_Lite_Segmentation_Jetson-Nano/archive/refs/heads/main.zip
$ unzip -j master.zip
Remove master.zip, LICENSE and README.md as they are no longer needed.
$ rm master.zip
$ rm README.md

Your MyDir folder must now look like this:
cat.jpg.mp4
deeplabv3_257_mv_gpu.tflite
TestUnet.cpb
Unet.cpp

---

Running the app.

Run TestTensorFlow_Lite.cpb with Code::Blocks.
You may need to adapt the specified library locations in TestTensorFlow_Lite.cpb to match your directory structure.

With the #define GPU_DELEGATE uncommented, the TensorFlow Lite will deploy GPU delegates, if you have, of course, the appropriate libraries compiled by bazel. Install GPU delegates

See the RPi 4 movie at: https://www.youtube.com/watch?v=Kh9DLMgCIIE

---