Tag: AIML

Making Sanskrit Accessible through AI-based Text Processing

Researchers at IIT Kharagpur are making Sanskrit accessible with their Artificial Intelligence-based system for processing Sanskrit texts

There has been a renewed interest in Sanskrit since the announcement of NEP 2020. Various academic institutions both at school education as well as higher education are adopting various approaches for improving the reach of the language through training programs, research and outreach initiatives. While various digital resources have improved the accessibility and use of world languages and even regional languages, Sanskrit presents unique challenges in automated computational processing. In addition to the sheer volume and diversity, both stylistic and chronological, found in these texts, the linguistic peculiarities expressed by the language, pose several challenges in making these works accessible to the world.  

Research group led by Dr. Pawan Goyal have developed a digital infrastructure for the efficient processing of Sanskrit texts, by effectively combining state-of-the-art machine learning techniques and traditional linguistic knowledge from Sanskrit. The proposed framework is based on Energy-based models and it enables the encoding of relevant linguistic information as constraints.

“Processing of Sanskrit texts poses several challenges owing to the high lexical productivity of the words, free word order in poetry, euphonic assimilation of sounds at the word boundaries and phonemic orthography followed in writing. Keeping these in mind, we proposed a generic graph-based framework that takes advantage of the free word order nature of the language. Further, we make use of linguistic insights from the traditional Sanskrit grammar for learning the feature function and applying the relevant constraints.” explained Dr. Goyal.

 

He further adds,  “Our proposed framework substantially reduces the training data requirements to as low as 10%, as compared to that of the neural state-of-the-art models. In all the Sanskrit-related tasks discussed in the work, we either achieve state-of-the-art results or ours is the only data-driven solution for those tasks.”

This work is accepted for publication in the Computational Linguistics journal published by the MIT Press. This work has been carried by research scholar Dr. Amrith Krishna [Tweet @krishnamrith12], currently, a postdoc at the University of Cambridge, supervised by Dr. Pawan Goyal. The paper currently addresses the tasks of word segmentation (संधि विच्छेद), morphological parsing (पद विश्लेषण), dependency parsing (कारक विश्लेषण) and poetry to prose conversion of Sanskrit text (अन्वय). The team is now actively collaborating with several external research groups to extend the application of the proposed system for automatic speech recognition and question-answering in Sanskrit.

The classical language has a rich literary tradition spanning more than 2 millennia that encapsulates the cultural ethos of this civilizational nation. Works in Sanskrit, numbering more than 30 million extant manuscripts, include extensive epics, subtle and intricate philosophical, mathematical, and scientific treatises, and rich literary, poetic, and dramatic texts. The proposed AI-based system, used in conjunction with interactive tools such as the Sanskrit Heritage reader, can aid the users in the easier analysis of these manuscripts with word-by-word analysis and translation, the relation between words, poetry to prose conversion, search and question answering, etc.

Graphic by Suman Sutradhar

For more information contact Prof. Pawan Goyal at pawang@cse.iitkgp.ac.in or Dr. Amrith Krishna [Face: amrith.krishna, Tweet @krishnamrith12]

Media Coverage:

 

Low-cost Solution for Advanced Manufacturing

IIT Kharagpur researchers develop a low-cost AI-based Industry 4.0 solution for real-time metrological inspection

In an initiative undertaken by the Centre of Excellence in Advanced Manufacturing Technology at IIT Kharagpur, researchers are on a mission to make AI & ML applications affordable for India’s industrial sector including MSMEs. They have developed an innovative system consisting of a low-cost imaging device and an  AI-enabled software for real-time metrological inspection. The developed system can be utilized in the production line to check the quality of the jobs and get instant results.

Explaining the benefit that MSMEs will draw from such an innovation, Prof. Surjya K Pal said, “The MSMEs mostly rely on manual inspection of the jobs produced in a batch where a few samples are randomly selected and checked manually. Accordingly, the entire batch is either rejected or accepted, which lacks effectiveness and increases the cost of production. The potential of the developed solution can be leveraged to inspection of each job in a batch, in real-time, and at a very minimum cost.”

This AI-enabled software will enhance the image quality captured by the low-cost camera at par with a high-quality camera output and process the image in real-time.  Further, it automates the acceptance or rejection of production jobs and delivers the outcome for real-time managerial insights.

With an accuracy close to 98%, the solution can measure the dimensional features of the job, and also it can inspect the presence of scratches, dents, etc., in a job. It takes only approximately 12 seconds for the analyses to be complete. The accuracy check and timeliness of the solution have been guaranteed by testing it on different types of jobs. Researchers are working further to reduce the time.

The research group included Prof. Surjya K Pal, Prof. Debashish Chakravarty, research scholar Debasish Mishra, technical staff Pravanjan Nayak, and intern Ayan Banerjee from Jalpaiguri Government Engineering College, who has filed a patent for the system and made it available for MSMEs to license the technology.

“Automation is one of the key aspects of digitalization and is also the need of the hour. The vision of manufacturers revolves around three things that are how fast can manufacturing be performed, how better, and how cheaper? The first highlights a higher rate of production, the second implies the use of innovation in manufacturing and the third aims at reducing the cost,” remarked Prof. Pal, the lead of Centre of Excellence in Advanced Manufacturing Technology.

While India has been making fast inroads in the domain of Artificial Intelligence and Machine Learning, it is the capital intensive and service industries which are able to put the applications to their vantage. The MSME sector, which employs more than 100 million people in the country contributing close to 30% of India’s GDP and 50% of export revenue, is still outside the ambit of industry-scale adoption of AI & ML.

“This is the gap which IIT Kharagpur aims to bridge while making low-cost Industry 4.0 solutions across all industrial sectors and accelerate the pace towards Atma Nirbhar Bharat,” he added.

Media Coverage:

Economic Times Hindustan Times Business Standard
ABP Education DataQuest Analytics India Magazine
NDTV The Week Outlook
CNBC TV18 Danik Jagran

Contact:

Project: Prof. Surjya K Pal, skpal@mech.iitkgp.ac.in;

Media: Shreyoshi Ghosh, shreyoshi@adm.iitkgp.ac.in

Follow IIT Kharagpur on social media: Facebook: @IITKgp Twitter: @IITKgp Instagram: @iit.kgp

AI4ICPS Setting Pace for India’s Transition to Smart Industry Hub

IIT Kharagpur is all set to transform the nation’s industrial sector with scalable application of Artificial Intelligence (AI) and Machine Learning (ML) in Cyber-Physical Systems (CPS) under the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS). The Institute has recently set up a special purpose vehicle, the IIT Kharagpur AI4ICPS I Hub Foundation for translation of academic research at the proof of concept level to industrially scalable products and processes, focused on AI interventions to CPS. 

“While research has been happening in AI&ML for decades, mostly limited to software with limited inclusion in consumer goods and daily used gadgets; graduating innovation to technology readiness for industrial production is crucial. The Hub is envisioned to be a unique establishment in the country to undertake a diverse spectrum of activities linking academic research to the industry through an inclusive co-working partnership, towards the common goal of unlocking the $500 billion contributions to India’s GDP expected from Data and AI by 2025,” remarked Principal Investigator Dr. Debdoot Sheet, Assistant Professor with the Department of Electrical Engineering and the Centre of Excellence in Artificial Intelligence. 

The mandate at AI4ICPS is to transition research from Technology Readiness Level (TRL) 3 to TRL 7, which means enabling the translation of academic research resulting in pre-/prototypes of AI&ML interventions in CPS to scalability ready products and processes for industry. The Institute received a grant of ₹ 170 crore from the Department of Science and Technology, Govt. of India under the NM-ICPS to set up the infrastructure and undertake the related research and training programs.

“We are also committed to promote fundamental research in TRL 1 to TRL 3 and foster the creation of startups to scale production for their off-shelf market availability, an activity beyond TRL 7 and ranging upto TRL 9,” added Dr. Sheet.

The Hub will further develop critical infrastructure for the Nation, an AI Cloud for CPS, a National Knowledge Portal for training human resources on the future-ready AI technology for CPS.

“The host of Digital Twins like tractors and farm machines, 3D printing technology, structural health and road traffic, wastewater engineering, energy-efficient buildings, next-gen wireless communication, human physiology and in-silico disease models are to name a few of the 32 technologies to be developed by the PEG at the Hub,” said Dr. Ashish Ranjan Hota, Assistant Professor with the Department of Electrical Engineering.

The R&D work will span over core research in algorithms, methods, explainability, trust and safety of AI&ML, enabling technologies like machine vision, speech, natural language processing (NLP), text analytics, augmented and virtual reality, technology for the internet of things (IoT), and also application of AI&ML technologies in a diverse range of industry verticals including (i) healthcare, (ii) precision agriculture and nutritional security, (iii) manufacturing, (iv) transportation, (v) environment and pollution, (vi) infrastructure and energy, (vii) education, (viii) judiciary and legal, and (ix) communication.

“IIT Kharagpur is a pioneer in AI education and research with more than a hundred faculty members working in theory and methods of AI and its applications across various disciplines. This Hub will create a multidimensional ecosystem to foster innovation and commercialization of AI&ML interventions to ICPS spanning across several sectors along with facilitation of new knowledge creation, upskilling human resources, entrepreneurship and job creation,” said Director, Prof. Virendra K Tewari. 

Recruitment is currently underway for leadership and engineering roles including CEO, CXOs, Engineering Managers to drive the pace at the Hub.

“AI4ICPS will have an in-house product engineering group, dedicated team of professionals to support technology commercialization and startup creation, and is looking to support more than 100 research fellows, postdoctoral researchers and early career academics over the next five years, to fulfil its mandate of promoting enterprises to add over 13,000 jobs in CPS,” said Dr. Ashish Ranjan Hota.

With research and new knowledge generation, the upskilling of human resources capable of meeting the demands of AI-powered industrial sectors is also on the chart.

“A mechanism of wide-scale training on AI technologies, not just for current graduates, but also for industry professionals is the need of the hour, to adapt to the rapidly evolving landscape of traditional industry imbibing AI into CPS. We need to train human resources across the spectrum of engineering, starting at diploma graduates, for the adoption of AI in CPS; and focus on building scalable technology with AI to enable wide-scale pedagogical reach out of AI for CPS,” added Prof. Sudeshna Sarkar from the Department of Computer Science & Engineering and Head of the Centre of Excellence in Artificial Intelligence.

Incubating startups to facilitate technology production and job creation will be another thrust area, with programs to support innovation promotion starting with student challenges, gradually maturing them with linked investments like ignition and seed support. The Hub programs on incubation and innovation acceleration will also offer business development training and handholding to technologists aspiring to be entrepreneurs in AI&ML technologies.  Technology Certification and Audit services for AI&ML intervention to ICPS, and their safety and verification shall also be provided at AI4ICPS as first of its kind offering within the country.

“The host of live testbeds of CPS made available via cloud will improve real time collaborations between academia and industry. We look forward to spin off cloud connected machines of the future, across sectors ranging from agriculture to heavy machinery,” added Dr. Rajendra Machavaram, Assistant Professor with the Department of Agricultural & Food Engineering. 

Prof. Surjya Pal from the Department of Mechanical Engineering & the Centre of Excellence in Advanced Manufacturing Technologies, who is associated with Hub mentioned his recent AI-based innovations of friction stir welding technology in collaboration with TCS and real-time metrological inspection which are significant steps in India’s journey towards Industry 4.0.

“Such AI-based innovations are the way forward for high-quality affordable products by means of minimizing production losses, production time and human intervention,” he said.

Over 80 Faculty members from diverse disciplines of the Institute in driving the Hub’s establishment. Their R&D work will have close collaboration with over 50 companies and PSUs who are engaging early on, scouting for potential technologies. Research collaborations are also being discussed with over 30 global research partners towards furthering the objective of a unifying effort to converge the world in building AI for industries of the future across sectors ranging from agriculture to heavy machinery. 

About AI 4 ICPS

AI4ICPS is a special purpose vehicle instituted for translation of academic research, generally available at proof of concept level to industrially scalable products and processes for Cyber-Physical Systems. Focus is on artificial intelligence (AI) and machine learning (ML) technologies applied across industry sectors like healthcare, precision agriculture and nutritional security, manufacturing, transportation, environment and pollution, infrastructure and energy, education, judiciary and legal, and communication. AI4ICPS engages in innovation translation through the Industry-Academia-Hub co-working product engineering group (PEG), and beyond it also promotes human resources and skill development for making available trained manpower for industry adopting such new technologies, promote entrepreneurship among the youth to facilitate new venture creation for technology production and new job creation, and collaborate with renowned groups to complement technology development through international cooperative research programs. The IIT KHARAGPUR AI4ICPS I HUB FOUNDATION is established as a Section 8 company with an assured investment of Rs. 170 crores by the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS) of the Department of Science and Technology, Government of India. The hub closely collaborates with 25 other hubs across the country for building a salubrious ecosystem for CPS industry in the country. 

For more information visit www.ai4icps.in

Media Coverage:

Hindustan Times EdEx Yahoo News
Analytics India India Blooms ABP Education
Outlook The Week NDTV
Dainik Jagran

Contact:

For project information: Prof. Debdoot Sheet, Assistant Professor, Department of Electrical Engineering and Centre of Excellence in Artificial Intelligence

Principal Investigator, AI 4 ICPS – Technology Innovation Hub on Artificial Intelligence and Machine Learning for Interdisciplinary Cyber-Physical Systems

Email: debdoot@ee.iitkgp.ac.in, project-director@ai4icps.in

For media queries:

Shreyoshi Ghosh, Executive Officer (Media & Communication), E: shreyoshi@adm.iitkgp.ac.in | shreyoshi.iitkgp@gmail.com

Follow us on social media: Facebook:@IITKgp | Twitter: @IITKgp | Instagram: @iit.kgp

 

Towards Farm Digitization

IIT Kharagpur Training Researchers in Application of Digital Technologies in Agriculture for bringing Digital Revolution in the Farm sector

In 2015, the Government of India had launched the Digital India programme to transition India’s public service sector into the digital space. Agriculture is the largest employment vertical with a GDP share of 14% holds a key place in this mission to digitization. Over the years, India has witnessed a substantial increase in yields, crop diversity due to mechanization and knowledge dissemination programmes. However, the steeping consumption due to population explosion and rising income has been a constant challenge despite the progress in the agricultural sector. While the Govt. of India has set up several programmes to achieve the dual solution of automation and digitization, technical institutions can facilitate the transformation. Towards this end, IIT Kharagpur is training the farming community and organizing outreach programmes. 

The Dept. of Agricultural and Food Engineering has launched an online course on “Application of Digital Technologies in Agriculture” under the Centre of Excellence for Digital Farming Solutions for Enhancing Productivity by Robots, Drones and AGVs project sponsored by NAHEP-CAAST, ICAR, New Delhi. The short-term course will focus on engaging smart technologies for enhancing productivity with minimal effort and cost. These would include Sensors, Drones, Robots, Artificial Intelligence and Machine Learning, Machine Vision Techniques, Computer-Aided Design and Advanced Digital Technologies application in Agriculture.

“Outreach projects in agricultural digitization are of paramount importance towards development and adoption of knowledge-intensive agriculture education. Globally researchers have been extending their support to precision farming and smart farming through innovations and data analytics. While our Prime Minister had announced a digital marketplace for the farming community, the technical institutions like ours also need to facilitate farming technology requirements of India. The recent calamities and their aftermath has shown why we need to be ready for bad monsoon or climate change or even return to organic farming ways while enhancing agricultural yields by means of technology,” said Prof. Virendra Kumar Tewari, Director, IIT Kharagpur and Principal Investigator of NAHEP-CAAST-IIT Kharagpur.

The NAHEP centre is integrated by three interdisciplinary research divisions such as Agribots, Agri-Drones and Agri-AGV’s based on four portfolios – 1. Climate-based Digital Knowledge Support Centre; 2. Seed/Seedling Processing and Nursery Automation Centre; 3. Smart Portable Machinery Centre; 4. Food Processing Automation Centre. It envisages the enhancement of quality and relevance of agricultural higher education to the agricultural university students.

“It is crucial to empower the human resources with digital technologies in the field of agricultural science and technology, particularly in less-explored domains. The penetration of such knowledge will drive further R&D and create a culture to understand and expedite the adoption of advanced technologies in agricultural practices,” said Prof. Rajendra Machavaram, joint Principal Investigator of NAHEP-CAAST-IIT Kharagpur.

The current course piloted with the postgraduate and doctoral students, faculty members, and scientists from the Vasantarao Marathwada Krishi Vidyapeeth, Maharashtra. However, it garnered significant interest among students and researchers from other institutions and also alumni of IIT Kharagpur. The online lectures for the course have been made available on the Google Meet Link: https://meet.google.com/mwg-cgbn-ddy

Media Coverage:

Careers 360       LiveMint         Outlook        NDTV         DataQuest        The Week

Contacts: 

Project Information: Prof. Rajendra Machavaram rajendra@agfe.iitkgp.ac.in;

Institute Related: Prof. B N Singh, registrar@hijli.iitkgp.ac.in; 

Media Outreach: Shreyoshi Ghosh, shreyoshi@adm.iitkgp.ac.in;

More News:: https://kgpchronicle.iitkgp.ac.in/; 

Follow IIT Kharagpur on Social Media: Facebook:@IIT.Kgp; Twitter: @IITKgp; Instagram: @iit.kgp

India’s Own AI Resource Platform

Indian Express     Outlook    NDTV    Careers360      Quint     Gadgets Now     The Week     The Statesman      NavBharat Times        Business Standard      Hindustan Times

IIT Kharagpur today announced that it will collaborate with Amazon Web Services, Inc. (AWS), an Amazon.com company, to develop a National Artificial Intelligence Resource Portal (NAIRP) platform powered by Amazon SageMaker to promote learning and development in Artificial Intelligence (AI) for all AI learners, practitioners, and researchers in India, and to make the search for AI learning resources simpler and more accessible.

IIT Kharagpur will use AWS Promotional Credits and Amazon SageMaker services to facilitate AI learning. In March 2019, IIT Kharagpur adopted the AWS Educate program to help students gain in-demand cloud computing skills, including hands-on AI experience on AWS Cloud.

NAIRP will provide access to computing facilities to users for AI learning and experimentation. Novice machine learning practitioners will be able to quickly search and get access to myriads of learning resources, and prototype the learned concepts into hands-on running modules without being held back by the lack of resources. In addition to this, the opportunity for joint research will facilitate the participation of students and researchers in contributing to open-source educational resources and workbook development for AI-related tasks.

“The contributions from AWS will enable us to provide cloud computing to all learners and reduce the high barrier to make training in AI more available,” said Professor Partha Pratim Chakrabarti, who conceptualized the development of NAIRP.

“There is already a great demand for professionals skilled in AI as well as resources, and this need will only increase in our society. The development of this platform will help boost the AI community in India,” said Professor Sudeshna Sarkar, Head, Centre for Artificial Intelligence at IIT Kharagpur.

Bratin Saha, Vice President AWS Machine Learning & Engines said, “AWS is delighted to work with IIT Kharagpur in developing the NAIRP platform powered by Amazon SageMaker, and we are excited about the opportunity to help make machine learning skills available to every developer. With more than 200 significant machine learning capabilities launched in the last two years, AWS has the broadest and deepest set of machine learning and AI services focused on solving some of the toughest challenges facing developers. We welcome the opportunity to work with IIT Kharagpur on some of those challenges.”

NAIRP was launched with initial funding from the Ministry of Human Resources Development, as part of the National Digital Library of India (NDLI) platform. NAIRP builds upon and expands NDLI to AI by indexing a wide variety of AI-specific resources such as learning materials, courses, workbooks, datasets, codes, tools and platforms. The resources will be classified by a comprehensive AI classification ontology and metadata standard which IIT Kharagpur is developing. AI resources from AWS as well as a variety of other sources such as Kaggle, GitHub, and National Programme on Technology Enhanced Learning (NPTEL) will also be indexed.

The National Artificial Intelligence Resource Portal will provide search, browse, learning and computation services. The support for learning will be in the form of several structured learning modules with instructional materials and workbooks, as well as through indexing of available courses and materials from NPTEL and other sources. Hands-on AI training will be facilitated through workbooks and the cloud. The system will start off using AWS Cloud, and will also be connected in the future with the in-house cloud developed at IIT Kharagpur.