Implementation of our ICIP 2020 paper "Lownet: privacy preserved Ultra-Low Resolution Posture Image Classification" In this project, we created LowNet architecture, which is suitable for low resolution image classification. We are releasing TIP38(Thermal Image Posture 38 class) yoga posture image dataset captured by infrared camera. We propose "Lownet" model with relu activation fucntions that have variable slopes.
We trained our model with different losses. Lownet model with our custom loss function yields better f1 score on our dataset.
Trained other SOTA models on our dataset and compared our LowNet model with these models. Our model performed better.
