Search Results for author:
Found 3 papers, 0 papers with code
Enhancing Cross-Category Learning in Recommendation Systems with Multi-Layer Embedding Training
At constant model quality, MLET allows embedding dimension, and model size, reduction by up to 16x, and 5. 8x on average, across the models.
Training with Multi-Layer Embeddings for Model Reduction
Increasing the dimension of embedding vectors improves model accuracy but comes at a high cost to model size.
Speech Enhancement using Adaptive Mean Median Deviation and EMD Technique
During the acquisition of the speech signal by the non-contact Speech Sensor (SS), the signal is degraded by severe colored noises which are non-linear and non-uniform in nature.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
