ML Research Wiki / Benchmarks / Spoken Language Understanding / Fluent Speech Commands

Fluent Speech Commands

Spoken Language Understanding Benchmark

Performance Over Time

📊 Showing 17 results | 📏 Metric: Accuracy (%)

Top Performing Models

Rank	Model	Paper	Accuracy (%)	Date	Code
1	UniverSLU 📚	UniverSLU: Universal Spoken Language Understanding for Diverse Tasks with Natural Language Instructions	99.80	2023-10-04	-
2	Finstreder (Conformer + AMT, character-based) 📚	Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models	99.80	2022-06-29	📦 Jaco-Assistant/Jaco-Master 📦 Jaco-Assistant/finstreder
3	Finstreder (Quartznet + AMT) 📚	Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models	99.70	2022-06-29	📦 Jaco-Assistant/Jaco-Master 📦 Jaco-Assistant/finstreder
4	textual-kd-slu 📚	Two-stage Textual Knowledge Distillation for End-to-End Spoken Language Understanding	99.70	2020-10-25	📦 clovaai/textual-kd-slu
5	Wav2Vec2.0-Classifier 📚	Integration of Pre-trained Networks with Continuous Token Interface for End-to-End Spoken Language Understanding	99.70	2021-04-15	-
6	E2E SLP two-step 📚	Speech-language Pre-training for End-to-end Spoken Language Understanding	99.70	2021-02-11	-
7	Wav2vec 2.0 SSL 📚	Do We Still Need Automatic Speech Recognition for Spoken Language Understanding?	99.60	2021-11-29	-
8	Finstreder (Conformer) 📚	Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models	99.50	2022-06-29	📦 Jaco-Assistant/Jaco-Master 📦 Jaco-Assistant/finstreder
9	AT-AT 📚	Exploring Transfer Learning For End-to-End Spoken Language Understanding	99.50	2020-12-15	-
10	BERT, AC Pretraining 📚	End-to-End Spoken Language Understanding for Generalized Voice Assistants	99.40	2021-06-16	-

All Papers (17)

UniverSLU: Universal Spoken Language Understanding for Diverse Tasks with Natural Language Instructions

2023

UniverSLU

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

2022

Finstreder (Conformer + AMT, character-based)

Jaco-Assistant/Jaco-Master Jaco-Assistant/finstreder

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

2022

Finstreder (Quartznet + AMT)

Jaco-Assistant/Jaco-Master Jaco-Assistant/finstreder

Two-stage Textual Knowledge Distillation for End-to-End Spoken Language Understanding

2020

textual-kd-slu

clovaai/textual-kd-slu

Integration of Pre-trained Networks with Continuous Token Interface for End-to-End Spoken Language Understanding

2021

Wav2Vec2.0-Classifier

Speech-language Pre-training for End-to-end Spoken Language Understanding

2021

E2E SLP two-step

Do We Still Need Automatic Speech Recognition for Spoken Language Understanding?

2021

Wav2vec 2.0 SSL

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

2022

Finstreder (Conformer)

Jaco-Assistant/Jaco-Master Jaco-Assistant/finstreder

Exploring Transfer Learning For End-to-End Spoken Language Understanding

2020

AT-AT

End-to-End Spoken Language Understanding for Generalized Voice Assistants

2021

BERT, AC Pretraining

Sequential End-to-End Intent and Slot Label Classification and Localization

2021

3D-CNN+LSTM+CE

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

2022

Finstreder (Quartznet)

Jaco-Assistant/Jaco-Master Jaco-Assistant/finstreder

Improving End-to-End Speech-to-Intent Classification with Reptile

2020

Reptile

FANS: Fusing ASR and NLU for on-device SLU

2021

FANS

Speech Model Pre-training for End-to-End Spoken Language Understanding

2019

Pooling classifier pre-trained using force-aligned phoneme and word labels on LibriSpeech

dscripka/openwakeword lorenlugosch/end-to-end-SLU

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

2022

Amazon Alexa

Jaco-Assistant/Jaco-Master Jaco-Assistant/finstreder

SpeechPrompt v2: Prompt Tuning for Speech Classification Tasks

2023

pGSLM+

Fluent Speech Commands

Performance Over Time

Edit Benchmark Results

Edit Result

Top Performing Models

All Papers (17)

UniverSLU: Universal Spoken Language Understanding for Diverse Tasks with Natural Language Instructions

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

Two-stage Textual Knowledge Distillation for End-to-End Spoken Language Understanding

Integration of Pre-trained Networks with Continuous Token Interface for End-to-End Spoken Language Understanding

Speech-language Pre-training for End-to-end Spoken Language Understanding

Do We Still Need Automatic Speech Recognition for Spoken Language Understanding?

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

Exploring Transfer Learning For End-to-End Spoken Language Understanding

End-to-End Spoken Language Understanding for Generalized Voice Assistants

Sequential End-to-End Intent and Slot Label Classification and Localization

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

Improving End-to-End Speech-to-Intent Classification with Reptile

FANS: Fusing ASR and NLU for on-device SLU

Speech Model Pre-training for End-to-End Spoken Language Understanding

Finstreder: Simple and fast Spoken Language Understanding with Finite State Transducers using modern Speech-to-Text models

SpeechPrompt v2: Prompt Tuning for Speech Classification Tasks

Model	Paper	Accuracy (%)	Date
UniverSLU	UniverSLU: Universal Spoken Language Understandin…	99.80	2023-10-04
Finstreder (Conformer + AMT, character-based)	Finstreder: Simple and fast Spoken Language Under…	99.80	2022-06-29
Finstreder (Quartznet + AMT)	Finstreder: Simple and fast Spoken Language Under…	99.70	2022-06-29
textual-kd-slu	Two-stage Textual Knowledge Distillation for End-…	99.70	2020-10-25
Wav2Vec2.0-Classifier	Integration of Pre-trained Networks with Continuo…	99.70	2021-04-15
E2E SLP two-step	Speech-language Pre-training for End-to-end Spoke…	99.70	2021-02-11
Wav2vec 2.0 SSL	Do We Still Need Automatic Speech Recognition for…	99.60	2021-11-29
Finstreder (Conformer)	Finstreder: Simple and fast Spoken Language Under…	99.50	2022-06-29
AT-AT	Exploring Transfer Learning For End-to-End Spoken…	99.50	2020-12-15
BERT, AC Pretraining	End-to-End Spoken Language Understanding for Gene…	99.40	2021-06-16
3D-CNN+LSTM+CE	Sequential End-to-End Intent and Slot Label Class…	99.30	2021-06-08
Finstreder (Quartznet)	Finstreder: Simple and fast Spoken Language Under…	99.20	2022-06-29
Reptile	Improving End-to-End Speech-to-Intent Classificat…	99.20	2020-08-05
FANS	FANS: Fusing ASR and NLU for on-device SLU	99.00	2021-10-31
Pooling classifier pre-trained using force-aligned phoneme and word labels on LibriSpeech	Speech Model Pre-training for End-to-End Spoken L…	98.80	2019-04-07
Amazon Alexa	Finstreder: Simple and fast Spoken Language Under…	98.70	2022-06-29
pGSLM+	SpeechPrompt v2: Prompt Tuning for Speech Classif…	98.20	2023-03-01