FVLLMONTI: The 3D Neural Network Compute Cube $(N^{2}C^{2})$ Concept for Efficient Transformer Architectures Towards Speech-to-Speech Translation.
Ian O'ConnorSara MannaaAlberto BosioBastien DeveautourDamien DeleruyelleTetiana ObukhovaCédric MarchandJens TrommerÇigdem ÇakirlarBruno Neckel WeslingThomas MikolajickOskar BaumgartnerMischa ThesbergDavid PirkerChristoph LenzZlatan StanojevicMarkus KarnerGuilhem LarrieuSylvain PelloquinKonstantinous MoustakasJonas MüllerGiovanni AnsaloniAlireza AmirshahiDavid AtienzaJean-Luc RouasLeila Ben LetaifaGeorgeta BordeallCharles BrazierChhandak MukherjeeMarina DengYifan WangMarc FrançoisHoussem RezguiReveil LucasCristell ManeuxPublished in: DATE (2024)
Keyphrases
- neural network
- speech recognition
- speech signal
- automatic speech recognition
- recognition engine
- broadcast news
- audio stream
- speech synthesis
- text to speech
- fuzzy logic
- efficient computation
- spontaneous speech
- speaker recognition
- speaker identification
- back propagation
- pattern recognition
- artificial intelligence
- audio visual
- bp neural network
- multilayer perceptron
- dialogue system
- spoken language
- parallel architectures
- machine translation
- fault diagnosis
- spoken dialogue systems
- data sources
- artificial neural networks
- vocal tract
- genetic algorithm