yoruba face

Yoruba Speech Dataset: A dataset for low-resource speech tasks.

  • 3 hours of audio recordings from a native yoruba speaker.
  • Can be used for low-resource speech tasks
  • Details of the project can be found on my github repo

yoruba face

Comparing Yoruba and French Speech Speech Models in Low-Resource Settings

  • CPC models is trained on 1 hour of yoruba and french audio with labels
  • Pretrained CPC model from Facebook CPC_audio library is then finetuned on 1hr of data.
  • Details of the project can be found on my github repo

exploring detectron2

Exploring Detectron2 for Object Detection Tasks

  • Instance segmentation, pose estimation, object tracking performed using the detectron2 library
  • Pretrained models finetuned on custom tasks
  • Details of the project can be found on my github repo
    Object tracking demo  

exploring detectron2

Cassava Disease Classification

  • Classify cassava plants as belonging to one of 4 various disease classes or healthy
  • Experimented with various CNN architectures for this task
  • Explore different methods of dealing with imbalanced datasets. I wrote a blog post of what worked for me
  • Details of the project can be found on my github repo

dna sequence

DNA Sequence Classification using Kernel Methods

  • Predict whether DNA sequence region is binding site using Kernel methods
  • Experimented with different kernels such as mismatch kernel, gaussian kernel, weighted-degree kernel
  • Details of the project can be found on my github repo

dog cat

Dog Cat Classifier

  • A simple ml pipeline for discriminating between dogs and cats
  • Experiment can be found on my github repo

dog cat

The Matching Game

  • A game written in React/Javascript to test attention and retention
  • Game hosted on netlify. You can play using the game link

Explore other experiments and projects in my github repository


© 2020 Sewade Ogun with help from Jekyll Bootstrap and Twitter Bootstrap
This work is licensed under a Creative Commons License