YSU (Your Simulated Universe) is a hybrid rendering engine combining C-based ray tracing, neural radiance fields, and a Vulkan real-time backend. Built from scratch, YSU explores the future of graphics through physics, computation, and intelligent light simulation.

A PyTorch implementation of NeRF trained on the NeRF Synthetic Blender Dataset (https://www.matthewtancik.com/nerf).

Neural Radiance Field from scratch

A PyTorch implementation of Neural Radiance Fields (NeRF) built from scratch.

Repository for records of studying Neural Radiance Fields from scratch

Neural Radiance Fields from scratch in one notebook with Tensorflow

This project presents a complete neural rendering and novel view synthesis system developed from scratch. It focuses on learning 3D scenes from multi-view images using optimizable volumetric radiance field models, such as Gaussian Splatting, to generate new perspectives from arbitrary camera viewpoints.

Implement a Neural Radiance Field from Scratch

Implemented Neural Radiance Field (NeRF) Pipeline from scratch

Neural Radiance Fields for 3D Scene Reconstruction — Built from scratch in PyTorch

Neural Radiance Fields for 3D Scene Reconstruction — Built from scratch in PyTorch

Neural Radiance Fields from scratch. Positional encoding, hierarchical volume rendering, view-dependent color, novel view synthesis.

Implementation of papers from scratch in the field of Neural radiance field to build my own understanding

Writing Neural Radiance Fields from scratch in python. Adapted from the assignment in the CS180 course at UC Berkley

Novel use of Neural Radiance Fields (NeRF) to create CT images from X-ray images. Implemented from scratch with PyTorch and SimpleITK.

Novel use of Neural Radiance Fields (NeRF) to create CT images from X-ray images. Implemented from scratch with JAX and SimpleITK.

Neural Radiance Fields (NeRF) from scratch in PyTorch — ray generation, volume rendering, hierarchical sampling, mesh extraction & 360° novel view synthesis

Implementing a Neural Radiance Field (NeRF) from scratch for learning purposes. Highly hardware constrained (~8GB VRAM). Low quality renders only.

A from-scratch implementation of Neural Radiance Fields. Implements volumetric rendering, positional encoding, and hierarchical sampling manually to demonstrate the core physics of neural rendering.

Gaussian Splatting, from scratch. Based on 3D Gaussian Splatting for Real-Time Radiance Field Rendering (https://arxiv.org/abs/2308.04079)

This repository contains a full, from-scratch implementation of Neural Radiance Fields (NeRF) using PyTorch. It covers core NeRF mechanisms including positional encoding, hierarchical volume sampling, and differentiable volume rendering, trained and validated on the Lego dataset to achieve high-fidelity novel view synthesis.

This repo is for the course project of EE5179: Deep Learning for Imaging at IIT Madras, under Prof. Kaushik Mitra, EE Dept. In this project, the Neural Radiance Field architecture was implemented for a custom dataset, created and processed from scratch.