5/10/24
Python, TensorFlow, JSON, CSV, Numpy, Pandas, Matplotlib, Seaborn
For my final Deep Learning project, my team decided to implement
this 2020 paper on NeRFs.
For those unfamiliar with NeRF, we trained a model on a collection of images of an object at different camera angles, as well as the
related camera angle information for each image. We then make the model generate the 3D object at a camera angle that the
model has not seen in the training data.
For this project, I worked mostly on making our dataset of images. I used Autodesk Maya and python scripts to create 600+ images
of a 3D model at different camera angles, and created a python script that would generate a JSON file of all camera information
for each image.
