Progress
This week, we worked towards figuring out the best classification algorithm and collecting more data to train our classifier on. Alvin and Ananya worked on implementing K-Nearest Neighbors, and they also extracted more features from the sound signal from the microphone. Tushita worked on training the ImageNet with more specific trash data as well as getting the Pi camera to take a picture and use the ImageNet on the Pi to classify the picture. We also worked on getting more precise readings from the capacitive sensor.
Next week, we plan on starting to build a full fledged sensor box and placing all of the circuitry into the plastic box. We also want to begin combining the results of the ImageNet classifier with the trinity sensor classifier.