Improving NeRF using Empty Space Skipping
Optimizing vanilla Neural Radiance Fields computation costs by incorporating State of the Art empty space skipping technique 'SparseLeap'.
Machine Learning
This section covers my projects in the field of Machine Learning and Deep Learning. The projects displayed here are all implemented using Machine Learning and Deep Learning techniques.
FaceSwap
Modifying the 'PRNet' model to swap faces.
SfMLearner
Adding Structural Similarity Index to the 'SfMLearner' model to improve monocular depth estimates.
HomographyNet
Training a deep Neural net to estimate Homographies for feature matching and creating Panoramic views.
Monocular pose estimation of cars using transfer learning and UNet architecture
Using the UNet architecture to estimate poses of cars in 3D using 2D monocular images.
Fundamentals of Artificial Intelligence and Deep Learning Frameworks
Implementing and analyzing classical Machine Learning algorithms and Neural Nets.