~/projects/speech-enhancement
Speech enhancement for a medical wearable
Recovered an abandoned speech enhancement prototype for a wearable device and compared it with two architectures of my own.
sheet
- role
- Research Data Scientist / Deep Learning Engineer, Robolabo (UPM)
- period
- 2024.05 → 2024.08
- context
- At Robolabo, UPM
- status
- archived
- code
- private
- stack
- Python, PyTorch
results
My transformer was the best of the three
Raised speech quality from 2.10 to 2.37 PESQ (+13%) on unseen audio with hospital noise
skills
- technical skills
- soft skills
- domains