Train neural network
Budget: $10 – $30 USD
Need to build and train a neural network model to localize the mice in the picture
The dataset is in the CSV file.
-Frist need to split the frames and the corresponding annotations to 70% for training and 30%
for evaluation.
● Train a model to predict mouse location and evaluate your model.
- Train a neural network model to predict the mouse location. The program should
be written in python.
- The inputs of the model are 7 joints’ coordinates.
- The mouse location ground truth is defined as the average of the corner
coordinates of the bounding box, i.e., center = ((x1+x2)/2, (y1+y2)/2).
- Loss function and evaluate metric are both L2 loss.
The dataset is in the CSV file.
-Frist need to split the frames and the corresponding annotations to 70% for training and 30%
for evaluation.
● Train a model to predict mouse location and evaluate your model.
- Train a neural network model to predict the mouse location. The program should
be written in python.
- The inputs of the model are 7 joints’ coordinates.
- The mouse location ground truth is defined as the average of the corner
coordinates of the bounding box, i.e., center = ((x1+x2)/2, (y1+y2)/2).
- Loss function and evaluate metric are both L2 loss.