scholarly journals Distilling weighted finite automata from arbitrary probabilistic models

2019 ◽  
Author(s):  
Ananda Theertha Suresh ◽  
Brian Roark ◽  
Michael Riley ◽  
Vlad Schogol
Keyword(s):  
Probabilistic Models ◽  
Finite Automata ◽  
Weighted Finite Automata

scholarly journals Approximating probabilistic models as weighted finite automata

2021 ◽  
pp. 1-36
Author(s):  
Ananda Theertha Suresh ◽  
Brian Roark ◽  
Michael Riley ◽  
Vlad Schogol
Keyword(s):  
Open Source Software ◽  
Probabilistic Models ◽  
Finite Automata ◽  
Source Model ◽  
Language Models ◽  
Leibler Divergence ◽  
Difference Of Convex ◽  
Difference Of Convex Optimization ◽  
N Gram ◽  
Weighted Finite Automata

Abstract Weighted finite automata (WFA) are often used to represent probabilistic models, such as n-gram language models, since, among other things, they are efficient for recognition tasks in time and space. The probabilistic source to be represented as a WFA, however, may come in many forms. Given a generic probabilistic model over sequences, we propose an algorithm to approximate it as a weighted finite automaton such that the Kullback-Leibler divergence between the source model and the WFA target model is minimized. The proposed algorithm involves a counting step and a difference of convex optimization step, both of which can be performed efficiently.We demonstrate the usefulness of our approach on various tasks, including distilling n-gram models from neural models, building compact language models, and building open-vocabulary character models. The algorithms used for these experiments are available in an open-source software library.

scholarly journals PATH-EQUIVALENT DEVELOPMENTS IN ACYCLIC WEIGHTED AUTOMATA

2007 ◽  
Vol 18 (04) ◽  
pp. 799-811
Author(s):  
MATHIEU GIRAUD ◽  
PHILLIPE VEBER ◽  
DOMINIQUE LAVENIER
Keyword(s):  
Surface Area ◽  
Finite Automata ◽  
Clock Frequency ◽  
Partial Removal ◽  
Frequency Constraints ◽  
Weighted Automata ◽  
Weighted Finite Automata ◽  
Special Case

Weighted finite automata (WFA) are used with FPGA accelerating hardware to scan large genomic banks. Hardwiring such automata raises surface area and clock frequency constraints, requiring efficient ∊-transitions-removal techniques. In this paper, we present bounds on the number of new transitions for the development of acyclic WFA, which is a special case of the ∊-transitions-removal problem. We introduce a new problem, a partial removal of ∊-transitions while accepting short chains of ∊-transitions.

Weighted Finite Automata Based Image Compression on Hadoop MapReduce Framework

2015 ◽  
Author(s):  
U.S.N. Raju ◽  
Irlanki Sandeep ◽  
Nattam Sai Karthik ◽  
Rayapudi Siva Praveen ◽  
Mayank Singh Sachan
Keyword(s):  
Image Compression ◽  
Finite Automata ◽  
Mapreduce Framework ◽  
Hadoop Mapreduce ◽  
Weighted Finite Automata

Similarity Enrichment in Image Compression through Weighted Finite Automata

2000 ◽  
pp. 447-456 ◽  
Author(s):  
Zhuhan Jiang ◽  
Bruce Litow ◽  
Olivier de Vel
Keyword(s):  
Image Compression ◽  
Finite Automata ◽  
Weighted Finite Automata

Languages Accepted by Integer Weighted Finite Automata

1999 ◽  
pp. 123-134 ◽  
Author(s):  
Vesa Halava ◽  
Tero Harju
Keyword(s):  
Finite Automata ◽  
Weighted Finite Automata

Fractal Coding Based Video Compression Using Weighted Finite Automata

2018 ◽  
Vol 9 (1) ◽  
pp. 115-133 ◽  
Author(s):  
Shailesh D. Kamble ◽  
Nileshsingh V. Thakur ◽  
Preeti R. Bajaj
Keyword(s):  
Video Compression ◽  
Signal To Noise Ratio ◽  
Finite Automata ◽  
Video Quality ◽  
Self Similarity ◽  
Fractal Coding ◽  
Time Compression ◽  
Practical Reality ◽  
Weighted Finite Automata ◽  
Evaluation Parameters

Main objective of the proposed work is to develop an approach for video coding based on Fractal coding using the weighted finite automata (WFA). The proposed work only focuses on reducing the encoding time as this is the basic limitation why the Fractal coding not becomes the practical reality. WFA is used for the coding as it behaves like the Fractal Coding (FC). WFA represents an image based on the idea of fractal that the image has self-similarity in itself. The plane WFA (applied on every frame), and Plane FC (applied on every frame) coding approaches are compared with each other. The experimentations are carried out on the standard uncompressed video databases, namely, Traffic, Paris, Bus, Akiyo, Mobile, Suzie etc. and on the recorded video, namely, Geometry and Circle. Developed approaches are compared on the basis of performance evaluation parameters, namely, encoding time, decoding time, compression ratio, compression percentage, bits per pixel and Peak Signal to Noise Ratio (PSNR). Though the initial number of states is 256 for every frame of all the types of videos, but we got the different number of states for different frames and it is quite obvious due to minimality of constructed WFA for respective frame. Based on the obtained results, it is observed that the number of states is more in videos namely, Traffic, Bus, Paris, Mobile, and Akiyo, therefore the reconstructed video quality is good in comparison with other videos namely, Circle, Suzie, and Geometry.

Quasi-Distances and Weighted Finite Automata

2017 ◽  
pp. 135-151
Author(s):  
Timothy Ng ◽  
David Rappaport ◽  
Kai Salomaa
Keyword(s):  
Finite Automata ◽  
Weighted Finite Automata
Sign in / Sign up

Export Citation Format

Share Document