ragibarnab Goto Github PK

amazon-fine-food-reviews-chat-assistant

Dipping my toes into the hype of Large Language Models. Created a assistant that answers questions about Amazon food reviews stored in an external data source. Uses retrieval from a vector database to augment the generation (RAG). Example Usage

cvd-binary-classifier-demo

Kaggle cats v. dogs binary classifier using convolutional neural networks in PyTorch.

dual-lidar-fusion

easy-self-driving-perception

"Simple, yet effective" approach to self-driving perception.

eeprom-temp-logger

A STM32 microcontroller project. Reads from a digital temperature sensor (TMP75B) with I2C and logs it into a serial EEPROM chip (25AA640A) using SPI.

flappy-rl

Pretty simple NEAT implementation for Pygame Flappy Bird from the PLE library.

incremental-sfm

Writing my own structure from motion programs in C++

nrf-demo

Configuring the popular nRF24L01 tranceiver module to communicate between two AVRs through the SPI protocol. Includes a handy C library for the ATmega328P.

octree-point-cloud-downsampler

A short project for downsampling point clouds using an octree.

pymar-i-o

Inspired by SethBling's "MarI/O - Machine Learning for Video Games" video, I implemented the NEAT algorithm on Python using OpenAI Gym Retro and NEAT-python libraries. It tries to tackle the 1st level of the original Super Mario Bros.

rae0924

raedl

A Django based website for myself.

ros2-darknet-pytorch-inference

ros2-lidar-object-detection

Easy and simple ROS 2 package to detect 3d boxes from lidar point clouds using PointPillars model implemented in PyTorch.

ros2-mmdetection3d-inference

ros2-realtime-semseg

ros2-semseg-lane-detection

Trying out semantic segmentation-based lane detection to generate lanes. WIP

sd-card-real-time-temp-logger

STM32 project that logs temperature and time into an SD card. Created to record the temperature of my room during power outages caused by snowstorms in Texas.

transfer-learning-snake-detection

An exercise in using transfer learning techniques and generating gradient CAMs (class activation maps).

ur5-poe-fkine

Forward kinematics exercise with product of exponential matrices for the UR5 robot.