audiomentations
A Python library for audio data augmentation. Useful for making audio ML models work well in the real world, not just in the lab.
- 2.3k
- Python
- MIT
- 7 mo ago
iver56
Machine learning, audio/music tech, computer vision, demoscene, web technology, games, startup. I mainly write Python. I also enjoy Go, C, Rust and JS.
Built from iver56's public GitHub profile, Trondheim, Norway. Not on Take My Repo yet, so there are no posts or updates here.
A Python library for audio data augmentation. Useful for making audio ML models work well in the real world, not just in the lab.
Fast audio data augmentation in PyTorch. Inspired by audiomentations. Useful for deep learning.
Evolving Artificial Neural Networks for Cross-Adaptive Audio Effects
Use a genetic algorithm to evolve an image by putting emojies on a canvas
A DNN that takes in x, y and outputs colors at that point to approximately "paint" an image. A toy problem.
The world's fastest Python package for calculating integrated loudness (LUFS) from audio data as NumPy arrays
Simple clap rhythm detection on Raspberry Pi using Csound and Python. Toy example.
Automatic temporally coherent video colorization designed to work well on animated cartoons and demos. End-to-end GAN.
Example of a basic python i18n setup with Gettext, Babel and Jinja2
A synthwave-style demo that ranked 1st in ARM DemoCompo in Trondheim, 2018. Designed to be run in Chrome on a Samsung Galaxy S9
Fast Python library for MP3 audio data augmentation (encode + decode for intentional audio quality degradation). Made for use in audiomentations.
Generate dataset with simple geometrical shapes, then train a RetinaNet on it and evaluate the result
A toy LSTM model that classifies gender based on first name. Presented at Trondheim Machine Learning Meetup in February 2018.
Toy project: Simulate beacon signal strengths and use them to estimate where the receiver is
Home automation: Light that turns on when somebody is standing still in front of the mirror
Sub-symbolic AI methods: Genetic Algorithms, Neural Networks and Emergent Behavior
Autoencoder toy problem: Convert 32x32 bitmap of a pong state to a 12-dimensional vector (that describes ball and paddle positions).
A convolutional neural network that can tell the time by looking at an image of an analog clock. A toy problem.