Multimodal Large Language Models with Fusion Low Rank Adaptation for Device Directed Speech Detection.
Shruti PalaskarOggi RudovicSameer DharurFlorian PesceGautam KrishnaAswin SivaramanJack BerkowitzAhmed Hussen AbdelazizSaurabh AdyaAhmed H. TewfikPublished in: CoRR (2024)
Keyphrases
- language model
- low rank
- speech recognition
- language modeling
- word error rate
- linear combination
- matrix factorization
- convex optimization
- n gram
- matrix completion
- low rank matrix
- missing data
- singular value decomposition
- audio visual
- spoken term detection
- probabilistic model
- rank minimization
- information retrieval
- retrieval model
- high dimensional data
- test collection
- query expansion
- semi supervised
- speech signal
- smoothing methods
- automatic speech recognition
- object detection
- high order
- trace norm
- data sets
- language models for information retrieval
- image processing
- recommender systems
- learning algorithm
- training data
- high dimensional
- vector space model
- small number
- image classification
- clustering method