1 . For this project we have tweaked Densenet-161 architecture
Dense Convolutional Network (DenseNet), connects each layer to every other layer in a feed-forward fashion. Whereas traditional convolutional networks with L layers have L connections - one between each layer and its subsequent layer - our network has L(L+1)/2 direct connections. For each layer, the feature-maps of all preceding layers are used as inputs, and its own feature-maps are used as inputs into all subsequent layers. DenseNets have several compelling advantages: they alleviate the vanishing-gradient problem, strengthen feature propagation, encourage feature reuse, and substantially reduce the number of parameters.
The 1-crop error rates on the imagenet dataset with the pretrained model are listed below.
Model structure Top-1 error Top-5 error
densenet121 : 25.35 : 7.83
densenet169 : 24.00 : 7.00
densenet201 : 22.80 : 6.43
densenet161 : 22.35 : 6.20
Download anaconda from here https://www.anaconda.com/distribution/#download-section
- Pytorch
conda install pytorch torchvision cudatoolkit=10.1 -c pytorch
- OpenCV
conda install -c conda-forge opencv
- Dataset of accident/non-accident images
https://drive.google.com/open?id=1o0D7vnGUZHS72is6n1jV1ge2BDfObzVi
- Pretrained Model binary file
https://drive.google.com/open?id=1AnJSogx65iyfIG0cSm5D15xfTGJzst8d
-
A proper php-language environment like xampp,remove htdocs folder and replace that with htdocs in this repo
-
Clone or Download this repo
git clone https://github.com/mad-skull/accident-detection.git
1.accident
2.Non-accident
Go to bash/cmd and type
python train.py
1. Traning set
2. Validation set
3.Number of corercts v/s epochs
Go to bash/cmd and type
python test.py
python evaluate.py
python livewebcam.py
The model reaches a classification accuracy of 86.00% accuracy on a randomly sampled test set, composed of 20% of the total amount of video sequences from our dataset. Will re-train this model when we have a good GPU and somre data .