We propose LIDSNet, a novel lightweight on-device intent detection model, which accurately predicts the message intent by utilizing a Deep Siamese Network.

Oct 6, 2021 · We further benchmark LIDSNet against fine-tuned. BERTs and show that our model is at least 41x lighter and 30x faster during inference than ...

To this end, we propose LIDSNet, a novel lightweight on-device intent detection model, which accurately predicts the message intent by utilizing a Deep Siamese ...

... Intent Detection and Slot Filling. 2020. 14. LIDSNet. 95.97. LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network. 2021. 15. Slot ...

Intent detection is a crucial task in any Natural Language Understanding (NLU) system and forms the foundation of a task-oriented dialogue system.

LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network. V Agarwal, SD Shivnikar, S Ghosh, H Arora, Y Saini. 20th IEEE International ...

LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network. V Agarwal, SD Shivnikar, S Ghosh, H Arora, Y Saini. 20th IEEE ...

Nov 1, 2021 · Releasing preprint of our work "LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network" to be presented at #IEEE ...

LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network ... Joint Slot Filling and Intent Detection via Capsule Neural Networks.

Title: LIDSNet: A Lightweight on-device Intent Detection model using Deep Siamese Network. Authors: Vibhav Agarwal, Sudeep Deepak Shivnikar, Sourav Ghosh ...