SincNet is a neural architecture for efficiently processing raw audio samples.

This project describes the necessary code to implement an EEG-based emotion recognition using SincNet [Ravanelli & Bengio 2018] including data from individuals diagnosed with Autism (ASD). For more details and data request send an email to the authors and contributors Juan Manuel Mayor Torres (juan.mayortorres@unitn.it) and Mirco Ravanelli (Mila)

Keras (tensorflow) implementation of SincNet (Mirco Ravanelli, Yoshua Bengio - https://github.com/mravanelli/SincNet)

Companion repository for the paper "A Comparison of Metric Learning Loss Functions for End-to-End Speaker Verification" published at SLSP 2020

Speaker change detection using SincNet and an LSTM/Transformer

The Additive Margin SincNet (AM-SincNet) is a new approach for speaker recognition problems which is based in the neural network architecture SincNet and the additive margin softmax (AM-Softmax) loss function. It uses the architecture of the SincNet, but with an improved AM-Softmax layer.

A PyTorch 1.0 implementation of the convolutions described in SincNet

Universal Adversarial Audio Perturbations

This implementation is based on the SincAlignNet model from the paper 'Frequency-Based Alignment of EEG and Audio Signals Using Contrastive Learning and SincNet for Auditory Attention Detection'.

Keras implementation of SincNet (https://github.com/mravanelli/SincNet, https://arxiv.org/abs/1808.00158)

Raw waveform adaptation with SincNet

An unofficial PyTorch implementation of MS-SincResNet [ICMR 2021] "MS-SincResNet: Joint learning of 1D and 2D kernels using multi-scale SincNet and ResNet for music genre classification"

An Implementation of SincNet using Tenorflow 2.x.

Tensorflow implementation of SincNet and its application to decode motor imagery from EEG data

No description available

My implementation of DENet, SincNet for bat call segmentation on the public dataset "Egyptian FruitBats"

End-to-End Music Emotion Recognition: Towards Language Sensitive Models

No description available

No description available

Automatic speaker gender recognition from raw input audio waveforms using a modified SincNet CNN arcitecture trained on the VoxCeleb Gender dataset using MatLab

[BIOSIG'2021] Curricular SincNet: Towards Robust Deep Speaker Recognition by Emphasizing Hard Samples in Latent Space.

No description available

Training and testing SincNet into TIMIT and Bengali_ASR dataset and model generation.

No description available

No description available

No description available

No description available

This is the implementation of the SincNet-based classifier, SincNet-R, which consists of three convolutional layers, and three deep neural network (DNN) layers for emotional EEG signals.

SincNet with Attention Mechanism using PyTorch

Classification using 1D complex raw-radar data