Welcome to our portal with new look.This portal has been redesigned to improve user friendliness
and appeal. If you have questions, comments or suggestions please give us your valuable
feedback
Indian Language Technology Proliferation & Deployment Centre
भारतीय भाषा प्रौद्योगिकी प्रसरण एवं विस्तारण केंद्र
Gaussian generative models have been shown to be equivalent to discriminative log-linear models under weak assumptions for acoustic modeling in speech recognition systems. In this paper, we note that the output layer of deep learning model consists of a first-order log-linear model, also known as logistic regression, which induces a set of homoscedastic distributions in the generative model space, resulting in linear decision boundaries.
Added on May 9, 2019
9
More Details
Contributed by : Individual
Product Type : Research Paper
License Type : Freeware
System Requirement :
Not Applicable
Author : Ankit Raj, Shakti P. Rath, Jithendra Vepa
Under the Indian Languages Corpora Initiative (ILCI) project initiated by the MeitY, Govt. of India, Jawaharlal Nehru University, New Delhi had collected corpus in Hindi as source language and translated it in Bodo as the target language. There are 70,000 sentences, including Health, Tourism, Agriculture and Entertainment domain in this corpus. This corpus has a unique sentence ID for each sentence, UTF-8 encoding, and text file format. The translated sentences have been POS tagged and Chunked properly. The chunking guideline used in this corpus creation, is provided in supporting document.
Under the Indian Languages Corpora Initiative (ILCI) project initiated by the MeitY, Govt. of India, Jawaharlal Nehru University, New Delhi had collected corpus in Hindi as source language and translated it in English as the target language. There are 70,000 sentences, including Health, Tourism, Agriculture and Entertainment domain in this corpus. This corpus has a unique sentence ID for each sentence, UTF-8 encoding, and text file format. The translated sentences have been POS tagged and Chunked properly. The chunking guideline is provided in supporting document.
Under the Indian Languages Corpora Initiative phase –II (ILCI Phase-II) project, initiated by the MeitY, Govt. of India, Jawaharlal Nehru University, New Delhi had collected Gujarati monolingual text corpus. There are approx. 30,000 sentences of general domain in this corpus. These sentences have been POS tagged and Chunked properly. The chunking guideline is provided in supporting document. This corpus has following features: unique ID, UTF-8 encoding, and text file format.
Under the Indian Languages Corpora Initiative (ILCI) project initiated by the MeitY, Govt. of India, Jawaharlal Nehru University, New Delhi had collected corpus in Hindi as source language and translated it in Gujarati as target language. There are 70,000 sentences including Health, Tourism, Agriculture and Entertainment domain in this corpus. Each sentence has a unique ID. The translated sentences have been POS tagged and Chunked properly. The chunking guideline is provided in supporting document.