Build an ensemble model using stacking classifier
Budget: ₹1,500 – ₹12,500 INR
We have an image dataset of 3090 images divided into test, train and validation folders. This is a custom dataset, collected from internet.
We need to do binary classification on these images, for which, we also have the target values.
The images have also been cropped to have just the facial regions and we have another folder of facial cropped images again divided into test, train and validation with their respective target values.
Results with the help of baseline algorithms have already been computed and fall in the range of 77-82% accuracy with both augmented and non-augmented data.
We now need to propose a new method through which we can demonstrate an increase in accuracy either through ensemble, or multi-input tensorflow functional api or few shot learning.
We need to do binary classification on these images, for which, we also have the target values.
The images have also been cropped to have just the facial regions and we have another folder of facial cropped images again divided into test, train and validation with their respective target values.
Results with the help of baseline algorithms have already been computed and fall in the range of 77-82% accuracy with both augmented and non-augmented data.
We now need to propose a new method through which we can demonstrate an increase in accuracy either through ensemble, or multi-input tensorflow functional api or few shot learning.