3D Gaussian Splatting (3DGS) is redefining the landscape of 3D computer graphics and vision — but here’s a twist: it achieves groundbreaking results without relying on any neural networks, not even a ...
Contrastive Learning – SimCLR and BYOL (With Code Example)
Supervised Learning has been dominant for years, but its reliance on labeled data—a costly and time-consuming resource—creates challenges, especially in areas like medical imaging. On the other hand, ...
The Annotated NeRF – Training on Custom Dataset from Scratch in Pytorch
In recent years, the field of 3D from multi-view has become one of the most popular topics in computer vision conferences, with a high number of submitted papers each year. A groundbreaking paper in ...
Introduction to Speech to Speech: Most Efficient Form of NLP
We often take out our phones and say, “Hey Siri, play Perfect by Ed Sheeran” or “Ok Google, set an alarm at 7.30 in the morning.” And the work is done on the flow by our phones! But have you ever ...
Training 3D U-Net for Brain Tumor Segmentation (BraTS2023-GLI) Challenge
3D U-Net, an efficient paradigm in medical segmentation, excels at analyzing 3D volumetric data, allowing it to capture a holistic view of brain scans. In many parts of the world, ...
DETR: Overview and Inference
In the groundbreaking paper “Attention is all you need”, Transformers architecture was introduced for sequence to sequence tasks in NLP. Models like Bert, GPT were built on the top of Transformers ...
