📌 Just launched a Multi-Class Fingerprint Image Dataset on Kaggle 1000 images | 5 classes (Arch, Tented Arch, Left/Right Loop, Whorl) Includes labeled CSVs – ideal for classification & biometrics. 🔗 [kaggle.com/datasets/shiva…]

soon will implement LeNet-5 architecture on this dataset from scratch in pytorch and will share the code and results

Advances in Computer Vision from MIT Open Courseware is available freely and is one of the fantastic courses.

I swear shud’ve started learning Ml in womb

📘 Implemented the LeNet-5 research paper from scratch using PyTorch. Includes full training pipeline, loss/accuracy plots, and original paper. 🔗 GitHub: github.com/Dgksv/Paper-Im…

📘 Implemented AlexNet from scratch in PyTorch Trained on CIFAR-10 → achieved 85.39% validation accuracy. Fully reproducible code + plots on GitHub 🔗 github.com/shivamd2502/Pa… #DeepLearning #PyTorch #ComputerVision #AI

replicated vgg16 from scratch using Pytorch on CIFAR-10 (tuning needed) github.com/shivamd2502/Pa…

honestly, impressed

implemented ResNet from scratch github.com/shivamd2502/Pa…

Published a blog on Byte Pair Encoding (BPE) — a key tokenization method behind models like GPT. The post combines research insights and GPT-driven explanations with step-by-step examples and frequency tables. medium.com/@shivamdapkeka…