test-time-augmentation

PyTorch extensions for fast R&D prototyping and Kaggle farming

Collection of awesome test-time (domain/batch/instance) adaptation methods

Image Test Time Augmentation with PyTorch!

Code Implementation for EmbedSeg, an Instance Segmentation Method for Microscopy Images

Test-Time Augmentation library for Pytorch

Efficient Test-Time Scaling for Small Vision-Language Models, official implementation of the ICLR'26 paper, test-time scaling via test-time augmentation

My modified version of YoloV5 training, cross-validation and inference with Pseudo Labelling pytorch pipelines used in GWD Kaggle Competition

Wheat detection using Faster RCNN

Engage in a semantic segmentation challenge for land cover description using multimodal remote sensing earth observation data, delving into real-world scenarios with a dataset comprising 70,000+ aerial imagery patches and 50,000 Sentinel-2 satellite acquisitions.

A curated collection of Test-Time Adaptation (TTA) papers, methods, and resources, updated monthly with the latest research.

Source code for the CIKM 2024 paper "Post-Training Embedding Enhancement for Long-Tail Recommendation."

My modified version of EfficientDet training, cross-validation and inference with Pseudo Labelling pytorch pipelines used in GWD Kaggle Competition

ARC-Test-Time-Training (ARC-TTT)

End-to-end HCR R&D: from HOG+XGBoost baseline to CNN hyperparameter optimization and Ensemble+TTA, achieving 93.16% accuracy with rigorous error analysis.

Object Detection and Bounding Box Prediction using YOLO5 and EfficientDet , Image Augmentations and Test Time Augmentations

Image recognition used to distinguish between bees and wasps in photographs

A modern implementation of ECCV 2018 paper: "Hierarchical Relational Networks for Group Activity Recognition and Retrieval".

Test time augmentation with Tensorflow keras models for segmentation tasks. This package also enables creation of keras layers for GPU acceleration

High-performance LPR system optimized for Indian license plates, achieving 97% character accuracy. Features a hybrid pipeline using YOLOv11 and Mamba-SSM (State Space Models) with built-in regex correction and Beam Search decoding.

This project explores workplace safety compliance in construction sites by evaluating a Convolutional Neural Network (CNN) for binary classification of images into "safe" (individuals wearing safety equipment) and "unsafe" (individuals not wearing safety equipment).