- 🔍 I am looking to connect with people in the fields of machine learning, computer vision, data science or data analytics. If my skills and experience align with your interests, feel free to contact me.
- 💡 My interests include data science, computer vision and research. Using machine learning and deep learning in various fields.
- 🧠 I have specific knowledge and experience in music computation and automatic music transcription.
- 💬 Feel free to ask me about automatic music transcription.
- 📫 Reach me at: hashimhussain4242@gmail.com
-
- Three regression ML models (polynomial regression, ridge regression and elastic net regression) to predict student performance. Including data cleaning and preparation, initial dataset analysis, model evaluation, dataset suitability analysis and observations.
-
NLP-Fake-News-Detection-with-Random-Forest-Classification-and-LSTMs (2024)
- Two fake news detector models, using ML (random forest classification) / DL (LSTMs) and BERT / TF-IDF feature extraction. See the evaluatory paper on the suitability of TF-IDF and BERT with the implemented models here.
-
VR-Museum-Data-Analytics (2022) (Internship)
- Analysis of VR data collected from a 3D museum environment to assess scientific relationships between head movements, eye movements and the virtual position of a person.
-
Data-Collection-and-Cleaning (2022)
- Pre-processing and cleaning three yearly survey datasets for visualisation. Using numpy, pandas, seaborn and matplotlib.
-
ML-Predicting-Age-From-Humour-Styles (2021)
- Three models to predict age from humour style using supervised learning methods - linear regression, polynomial regression and KNN. Read the paper here.
-
audio-visual-music-transcription (2024) / audio-visual-piano-transcription (2024) (Most starred ⭐)
- The first publically avaliable joint automatic audio-visual piano transcription model. For videos from an overhead angle. Achieves near state-of-the-art results
- Including a visual model with keyboard detection, hand detection and key detection. Where the notes covered by hands for each frame are passed as input to an audio model
-
Computer-Vision-Style-Transfer (2024)
- A movie-game style transfer model, using a CycleGAN model. Improved with human detection and human classification to further enhance the look of humans within frames
For my complete portfolio, including: software engineering, further deep learning, reinforcement learning, computer graphics, security, game development and more, please explore the additional public repositories on my GitHub profile.
