Rezultati pretraživanja
  1. prije 2 sata

    Sony workshop

    Three papers got accepted .
    1 proslijeđeni tweet 3 korisnika označavaju da im se sviđa
  2. prije 10 sati

    Our recently accepted paper at on neural multichannel speech enhancement is available on arxiv: , which shows connection between channel attention mechanism and classical beamforming!

    3 proslijeđena tweeta 5 korisnika označava da im se sviđa
  3. 1. velj

    Quite happy to share that our work on combining SincNet with the CTC+Attention training procedure for E2E ASR from the raw waveform has been accepted at . Paper: Code:

    Interested in E2E speech recognition from the raw waveform? We propose to combine SincNet and the Joint CTC-attention training to achieve this goal! (4.7% of WER on WSJ) :D
    8 proslijeđenih tweetova 18 korisnika označava da im se sviđa
  4. 31. sij

    Congratulations to Benjamin Lentz, Anil Nagathil, Johannes Gauer, and Rainer Martin for their accepted paper on HARMONIC/PERCUSSIVE SOUND SEPARATION AND SPECTRAL COMPLEXITY REDUCTION OF MUSIC SIGNALS FOR COCHLEAR IMPLANT LISTENERS

    3 korisnika označavaju da im se sviđa
  5. 31. sij

    Clotho audio captioning dataset will be presented in IEEE ! Join us to learn more about the novel and exciting task of audio captioning! Clotho is freely available:

    1 proslijeđeni tweet 2 korisnika označavaju da im se sviđa
  6. 30. sij

    Congratulations to Stefan Thaleiser and Gerald Enzner for their accepted paper "A COMPUTATIONALLY LIGHT ALGORITHM FOR BAYESIAN SPEECH ENHANCEMENT WITH SNR MARGINALIZATION"

    3 korisnika označavaju da im se sviđa
  7. 29. sij

    Details about the proposed method

    1 proslijeđeni tweet 2 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  8. 29. sij

    Work on training models from generated images has been accepted at

    This dataset does not exist: training models from generated images -- new work with V. Besnier and team Recent GANs, e.g. BigGAN, output increasingly high-quality images. What if we trained classifiers over fake images only, an initial aim of GANs? 1/
    Prikaži ovu nit
    1 proslijeđeni tweet 4 korisnika označavaju da im se sviđa
  9. 29. sij

    I am co-author of 3 papers but not presenting but as it is in Barcelona I though ok, let's go. Early rate €866.07 makes me think twice, not very accessible for Spanish researchers :-(

    1 proslijeđeni tweet 18 korisnika označava da im se sviđa
  10. 29. sij

    Glad to announce a new work with my student Sudhakar Kumawat in 2020 - "Depthwise-STFT based separable Convolutional Neural Networks". Arxiv version:

    4 korisnika označavaju da im se sviđa
  11. 28. sij
    Odgovor korisniku/ci

    Awesome - we have quite a few papers from our team as well. See you at

    1 reply 3 korisnika označavaju da im se sviđa
  12. 28. sij

    I'm going to IEEE International Conference on Acoustics, Speech & Signal Processing (ICASSP 2020), May 4-8 in Barcelona, Spain. Join me!

    1 reply 2 korisnika označavaju da im se sviđa
  13. 28. sij

    Our paper "Multi-scale Octave Convolutions for Robust Speech Recognition" has been accepted to ! See you in Barcelona :)

    5 korisnika označava da im se sviđa
  14. 28. sij

    I'm a big fan of self-supervised deep learning. I think it will be widely adopted in the future in many fields, including speech. Maybe this will contribute to "make pre-training great again"! hahaha

    1 reply 4 proslijeđena tweeta 15 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  15. 28. sij

    PASE+ is an improved version of PASE for robust speech recognition. We employ an online speech distortion module, a revised encoder that better learns short- and long-term speech dynamics, and a refined the set of workers that encourage better cooperation.

    1 proslijeđeni tweet 8 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  16. 28. sij

    I'm happy to announce our latest work on self-supervised learning for . PASE+ is based on a multi-task approach useful for recognition. It will be presented at . paper: code: @Mila

    7 replies 82 proslijeđena tweeta 302 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  17. 28. sij

    6 papers of were accepted to . You can see the papers: Congrats to all Clova AI authors!

    1 reply 13 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  18. 28. sij

    Our paper related to semantic embeddings of visually grounded speech has been accepted to . Content coming soon.

    5 korisnika označava da im se sviđa
  19. 27. sij

    Accepted as oral for

    Prikaži ovu nit
    Prikaži ovu nit
  20. 27. sij
    Odgovor korisniku/ci

    Hola, there is your unroll: Thread by : I am thrilled to announce our paper “Feedback Recurrent AutoEncoder” was accepted at !… . See you soon. 🤖

    2 korisnika označavaju da im se sviđa

Nema rezultata.

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.