Bad accurancy , re-trained the SSD network for people detection #1172

sylia · August 2, 2021, 8:12am

Hello,

I re-trained the SSD network for people detection, and when I did an evaluation with eval-ssd.py I got only 41% accuracy,
batch size = 4,
epoch = 108,
Is there a way to have better accuracy in detection of people?

the retraining was performed on a PC ( GPU NVIDIA GeForce RTX 2060 )
Thank you

AastaLLL · August 2, 2021, 9:11am

Hi,

This is the Jetson Nano board.
Do you use jetson-inference for re-training?
If not, would you mind filing a topic on a desktop-related board to get the support?

Thanks.

sylia · August 2, 2021, 9:54am

I use jetson inference for training

github.com

dusty-nv/jetson-inference/blob/master/docs/pytorch-ssd.md

<img src="https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/deep-vision-header.jpg" width="100%">
<p align="right"><sup><a href="pytorch-collect.md">Back</a> | <a href="pytorch-collect-detection.md">Next</a> | </sup><a href="../README.md#hello-ai-world"><sup>Contents</sup></a>
<br/>
<sup>Transfer Learning - Object Detection</sup></s></p>

# Re-training SSD-Mobilenet

Next, we'll train our own SSD-Mobilenet object detection model using PyTorch and the [Open Images](https://storage.googleapis.com/openimages/web/visualizer/index.html?set=train&type=detection&c=%2Fm%2F06l9r) dataset.  SSD-Mobilenet is a popular network architecture for realtime object detection on mobile and embedded devices that combines the [SSD-300](https://arxiv.org/abs/1512.02325) Single-Shot MultiBox Detector with a [Mobilenet](https://arxiv.org/abs/1704.04861) backbone.  

<a href="https://arxiv.org/abs/1512.02325"><img src="https://github.com/dusty-nv/jetson-inference/raw/dev/docs/images/pytorch-ssd-mobilenet.jpg"></a>

In the example below, we'll train a custom detection model that locates 8 different varieties of fruit, although you are welcome to pick from any of the [600 classes](https://github.com/dusty-nv/pytorch-ssd/blob/master/open_images_classes.txt) in the Open Images dataset to train your model on.  You can visually browse the dataset [here](https://storage.googleapis.com/openimages/web/visualizer/index.html?set=train&type=detection).

<img src="https://github.com/dusty-nv/jetson-inference/raw/dev/docs/images/pytorch-fruit.jpg">

To get started, first make sure that you have [JetPack 4.4](https://developer.nvidia.com/embedded/jetpack) or newer and [PyTorch installed](pytorch-transfer-learning.md#installing-pytorch) for **Python 3.6** on your Jetson.  JetPack 4.4 includes TensorRT 7.1, which is the minimum TensorRT version that supports loading SSD-Mobilenet via ONNX.  And the PyTorch training scripts used for training SSD-Mobilenet are for Python3, so PyTorch should be installed for Python 3.6.

## Setup

> **note:** first make sure that you have [JetPack 4.4](https://developer.nvidia.com/embedded/jetpack) or newer on your Jetson and [PyTorch installed](pytorch-transfer-learning.md#installing-pytorch) for **Python 3.6**

This file has been truncated. show original

Thanks

dusty_nv · August 2, 2021, 2:28pm

Please see my follow-up from https://github.com/dusty-nv/jetson-inference/issues/1172

system · October 10, 2021, 3:41am

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.