Name: Linh T. Duong
Type: User
Company: KTH Royal Institute of Technology
Bio: Obtain a PhD in medical microbiology from the University of Greifswald, Germany. Interested in Microbiology, Molecular biology, Epidemiology, & Machine Learning
Twitter: DuongTuanLinh1
Location: Stockholm - Sweden
Blog: https://scholar.google.com/citations?user=aZKRy1oAAAAJ&hl=en
Linh T. Duong's Projects
Automatic detection of foreign objects on chest X-rays
A vision-based tool for transport system flow analysis using Faster R-CNN. This system is able to detect different types of objects, e.g., cars, buses, pedestrians, and classify as well as count them in transport videos with a real-time speech.
Object Detection and Image Segmentation with Detectron2
Most popular metrics used to evaluate object detection algorithms.
Some experiments with object detection in PyTorch
OCaml bindings for PyTorch
A simple implementation of ocrmypdf and tesseract with flask for hosting to a server as an API. The code was written on CentOS7. This code works on linux only as ocrmypdf library does not have support on windows because of missing leptonica dll. For windows consider https://github.com/lakshay1296/OCR_Conversion_JPEG2PDF. This is image to ocr pdf conversion.
Python-based tools for document analysis and OCR
OHEM support for Fast R-CNN
Download and visualize single or multiple classes from the huge Open Images v4 dataset
ONNX.js tutorial for heartbeat by Fritz.ai
Open Source Computer Vision Library
Comparison benchmarks between public force fields and Open Force Field Initiative force fields
An Open-Source Toolkit for Heterogeneous Information Network Embedding (HINE)
High level API for using machine learning models in OpenMM simulations
Open Source Neural Machine Translation in PyTorch
Attention Unet model with post process for retina optic disc segmention
PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice (『飞桨』核心框架,深度学习&机器学习高性能单机、分布式训练和跨平台部署)
Object detection and instance segmentation toolkit based on PaddlePaddle.
Awesome OCR toolkits based on PaddlePaddle (3.5M practical ultra lightweight OCR system, support training and deployment among server, mobile, embedded and IoT devices)
Deploy Pytorch models to production via panini
The released code for the paper: Pooling Architecture Search for Graph Classification, in CIKM 2021.
Pelee: A Real-Time Object Detection System on Mobile Devices
API code for fastai Penguin Inference (deployed as AWS Lambda function)
Perception algorithms for Self-driving car; Lane Line Finding, Vehicle Detection, Traffic Sign Classification algorithm.
NCBI Prokaryotic Genome Annotation Pipeline
Sentiment classification for Vietnamese text using PhoBert
Plant Disease Detector Web Application
Implementation of Principal Neighbourhood Aggregation for Graph Neural Networks in PyTorch, DGL and PyTorch Geometric
BEST SCORE ON KAGGLE SO FAR , EVEN BETTER THAN THE KAGGLE TEAM MEMBER WHO DID BEST SO FAR. The project is about diagnosing pneumonia from XRay images of lungs of a person using self laid convolutional neural network and tranfer learning via inceptionV3. The images were of size greater than 1000 pixels per dimension and the total dataset was tagged large and had a space of 1GB+ . My work includes self laid neural network which was repeatedly tuned for one of the best hyperparameters and used variety of utility function of keras like callbacks for learning rate and checkpointing. Could have augmented the image data for even better modelling but was short of RAM on kaggle kernel. Other metrics like precision , recall and f1 score using confusion matrix were taken off special care. The other part included a brief introduction of transfer learning via InceptionV3 and was tuned entirely rather than partially after loading the inceptionv3 weights for the maximum achieved accuracy on kaggle till date. This achieved even a higher precision than before.