The architecture of a system for full-text search by speech data based on a global search index
Annotation
This paper presents the architecture of a system for full-text search by speech data based on a global search index that combines information about all speech recordings in the archive. The architecture includes two independent blocks: an indexing block, and a block for building and performing a search query. In order to process speech recordings, it uses an automatic speech recognition system (ASR) with a linguistic decoder based on weighted finite-state transducers framework (WFST), which generates word lattices. Lattices are sequentially converted to confusion networks and inverse indexes. It allows taking into account all the word hypotheses generated during decoding. The proposed solution expands the applicability of speech analytics systems for those cases when the word error rate is high, such as the processing of speech recordings collected under difficult acoustic conditions or in low-resource languages.
Keywords
Постоянный URL
Articles in current issue
- Features of images of water, ice, snow, objects and a human formed by a hybrid television camera in the near-infrared range
- Analyzing periodical textured silicon solar cells by the TCAD modeling
- Scintillation gamma radiation sensors based on solid-state photomultipliers in wireless industrial internet networks
- Improving the quality of network management of technological processes
- Geometric approach to the solution of the Dubins car problem in the formation of program trajectories
- Drift of two-dimensional vacancy islands on the Si(100) surface under electromigration conditions
- A study of the photocatalytic properties of chitosan-TiO2 composites for pyrene decomposition
- Kinetics of transformation of the atomic step bunches shape under electromigration conditions on the Si(001) surface
- Abnormal diffusion profile of adatoms on extremely wide terraces of the Si(111) surface
- An experimental methodology for assessing the probability and danger of network attacks in automated systems
- A meta-feature selection method based on the Auto-sklearn framework
- Automatic construction of the dialog tree based on unmarked text corpora in Russian
- Generic programming with combinators and objects
- Machine learning of the Bayesian belief network as a tool for evaluating the process frequency on social network data
- Software restructuring models for object oriented programming languages using the fuzzy based clustering algorithm
- The concept of managing the network structure of intelligent devices in the digital transformation of the energy industry
- Protecting facial images from recognition on social media: solution methods and their perspective
- Redundant models of testable distributed real-time computing systems
- A study of the influence of the base thickness on photoelectric parameters of silicon solar cells with the new TCAD algorithms
- A balanced algorithm of the hybrid large-particle method and its verification on some test problems
- Assessment of cerebral circulation through an intact skull using imaging photoplethysmography