Welcome,
to my digital
creative space

I'm Febryan Putra Kartika, Here, you'll find a curated collection of my projects, reflecting my dedication to machine learning.

About Me

I am from Indonesia, got my bachelor of science degree from National Taipei University of Technology, Taiwan. Very enthusiast about technology and machine learning, always try to keep up for the latest state-of-the-art technologies! Currently learning about machine learning albeit in the realm of NLP and CV, specifically I am interested to contribute to the community by improving healthcare quality through AI implementations.

Feel free to explore my portfolio and get a glimpse into the diverse projects I've been fortunate to work on. Each projects is a testament to my commitment to understand, implement and contribute my passion for machine learning. If you're curious about collaborations or have a project in mind, I'd love to hear from you

2018

Started learning python, data structures and algorithm

2019

Got introduced into machine learning and it's applications

2020

Created first ever CNN model for facial emotion recognition

2021

Dived deeper to understand more about neural networks, deep learning and theories behind state-of-the-art models

2022

Experimented with Transformer models to generate captions based on a given image (Bachelor's thesis)

2023

Conducted some experiments with speech recognition and NLP to help with early detection for medical disorders

Projects

Health Intent Classification with Speech Recognition

An experiment intended to create a model that could recognize what symptoms we had, using a speech recognition model to convert speech to text

Languages

Python
PyTorch
NLTK

Code

Source

ViT: Vision Transformer for Early Dementia Detection

An attempt to manually create a Vision-Transformer model from scratch, this model is based on the paper An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale, in this attempt I recreated the ViT-B/16 model.

Languages

Python
PyTorch

Code

Source

FoodVision

Used transfer learning method creating a custom Vision-Transformer model that could detect a specific food classes, this model is trained using 20% dataset from PyTorch Food101 dataset, and the model is hosted on HuggingFace Spaces

Languages

Python
PyTorch
HuggingFace

Code

Source

COVID-19 Detection

The purpose of this project is to compare models between TinyVGG and EfficientNetB0 on which model is better to detect COVID-19 from X-Ray images

Languages

Python
PyTorch
Transfer Learning

Code

Source

Image Caption Generator (Bachelor's Thesis)

Created an image to caption generator with Transfer Learning and Deep Learning from an Inception V3 model combined into a transformer encoder and decoder model

Languages

Python
Deep Learning
Transfer Learning
Tensorflow

Code

Source

Emotion Recognition

First project that is related to data science and machine learning, creating a CNN and training it to be able to accurately detect emotions including Happy, Sad, Angry, Scared, and Shocked

Languages

Python
Tensorflow

Code

Source

"Crinema" Online Movie Rating System

An experiment to create a functional rating system based on movies, the reason for this project is for movie lovers who at times have troubles to watch good movies, and this system will create a trust-worthy rating system that every movie lovers can rate and view regarding what movies they might find themselves suitable with

Languages

HTML
CSS
JavaScript
Git
Firebase

Code

Source

Minesweeper

An old school windows game when the internet was not a thing, this game keeps lots of people company when they are bored!

Languages

React
SCSS
TypeScript

Code

Source

Experiences

Worked on several machine learning projects, including object detection, intent classifications, text generation, medical detection and paper replications

Machine Learning
Familiar with Machine Learning algorithms (Linear and Logistic Regression, Decision Tree, Random Forest, Naive Bayes, SVM, KNN) and library such as PyTorch and Tensorflow
Healthcare
Interested in how healthcare quality could be improved with the help of innovative machine learning solutions and AI powered tools from early detection, diagnosis, treatment, and monitoring of diseases.
Natural Language Processing
Experienced with NLP theory such as Tokenizations, Vectorization, Bag of Words model, Lemmatization and Stemming, markov models also experimented with transformer models and how to fine-tune LLM such as BERT, T5, etc. Currently understanding more about NLP and Speech Technologies to facilitate more natural conversations between machines and humans, ultimately bridging the gap, enabling seamless and natural interactions between machines and users.
Computer Vision
Proficient in CV, including image preprocessing, deep learning architectures like CNNs, and implementing object detection and image segmentation algorithms. Utilized state-of-the-art models such as YOLO, and Vision Transformers (ViT) for various tasks. Fine-tuned pre-trained models such as ResNet and VGG. Interested in learning more about CV for Biomedical Image Analysis