Advanced Training Program on AI/ML

iHub-Data, IIIT Hyderabad plans to conduct a two-week executive training program on AI/ML for Scientists from DRDO to boost the R&D efforts of strategic research institutions in India. The program extended a brief theoretical overview of contemporary problem-solving paradigm involving AIML techniques and enabling practice in terms of:

(i) use of advanced AIML tools/libraries

(ii) use of public resources (datasets, trained models), and

(iii) training of complex ML models and their evaluation.

This short-term Executive Training program is intended to update the scientists of Indian Defence Sector Undertakings working in strategic sectors about the recent innovations taking shape around various AI/ML tools, techniques and architectural frameworks, from an end-to-end perspective.

Invited participants from select institutions who have undergone the 2-week Executive Training Program on AI/ML at IIITH will form the target audience for this course. Probable dates of commencement for this course is during second and third week of December 2024.

Schedule of the Sessions

09 December 2024

• 9 a.m. – 10 a.m. │ AIML Recap 1 (L) │ Dr Monalisa Patra
• 10 a.m. -11:15 a.m. │ Computer Vision through Ages (L) │ Prof P J Narayanan
• 11:30 a.m. to 1 p.m. │ Internet of Things – Introduction and Applications (L) │ Dr Sachin Chaudhari
• 2 p.m. – 5 p.m. │ Internet of Things – Introduction and Applications (T) │ Dr Sachin Chaudhari

10 December 2024

• 9 a.m. – 10 a.m. │ AIML Recap 2 (L) │ Dr Monalisa Patra
• 10 a.m. – 1 p.m. │ Explainability in ML Systems (L+T) │ Thrupthi Ann John
• 2 p.m. – 3:30 p.m. │ SLAM (Simultaneous Localization and Mapping) Pipeline and Beyond (L) │ Prof Madhava Krishna
• 3:30 p.m. – 5 p.m. │ AIML Recap 3 (L) │ Dr Monalisa Patra

11 December 2024

• 9 a.m. – 10 a.m. │ AIML Recap 4 (L)│ Dr Monalisa Patra
• 10 a.m. – 1 p.m. │ Motion Planning for Multi-UAV (Uncrewed Aerial Vehicles) Systems (L+T) │ Dr Harikumar K
• 2 p.m. – 5 p.m. │ AIML Recap 5 (L) │ Dr Monalisa Patra

12 December 2024

• 9 a.m. – 10 a.m. │ AIML Recap 6 (L) │ Dr Monalisa Patra
• 10 a.m. – 1 p.m. │ Large Language Models and Applications (L+T) │ Dr Vasudeva Varma
• 2 p.m. – 5 p.m. │ Vision Transformers and Applications (L+T)│ Dr Girish Varma

13 December 2024

• 9 a.m. – 10 a.m. │ Flexible Pressure Sensors and Their Applications (L) │ Dr Aftab Hussain
• 10 a.m. – 1 p.m. │ LLMs and Agents (L+T) │ Dr Vasudeva Varma
• 2 p.m. – 5 p.m. │ Detection and Tracking (L+T) │ Prof Vineet Gandhi

16 December 2024

• 9 a.m. – 10 a.m. │ Practical introduction to GenAI applications (L+T) │ Thrupthi Ann John
• 10 a.m. – 1 p.m. │ Computational Imaging (L+T) │ Prof Anoop Namboodiri
• 2 p.m. – 5 p.m. │ Discrete Communication (TTS) (L+T) │ Prof Vineet Gandhi

17 December 2024

• 9 a.m. – 10 a.m. │ CNNs (T) │ Seshadri Mazumder
• 10 a.m. – 1 p.m. │ AI Safety: Interpretability, Bias, Machine Unlearning, Robustness (L+T)│ Prof Ponnurangam Kumaraguru
• 2 p.m. – 5 p.m. │ AI/ML in CogSci Applications (L+T) │ Dr Vishnu Sreekumar

18 December 2024

• 9 a.m. – 10 a.m. │ RNNs and Attention (T) │ Seshadri Mazumder
• 10 a.m. – 1 p.m. │ Event-driven Data Pipelines for AI-enabled Systems (L+T)│ Dr Karthik V
• 2 p.m. – 5 p.m. │ Multimodal Large Language Models (L+T) │ Dr Makarand T

19 December 2024

• 9 a.m. – 10 a.m. │ Transformers (T) │ Seshadri Mazumder
• 10 a.m. – 1 p.m. │ Transformers: Intuition, Applications and Limitations (L +D)│ Prof C V Jawahar
• 2:30 p.m. – 4:30 p.m. │ Engineering Sustainable AI Systems (L+T) │ Dr Karthik V

20 December 2024

• 9 a.m. – 10 a.m. │ Feedback and Way Forward │ Dr C K Raju
• 10 a.m. – 11 a.m. │ Association with IIITH │ Prof Yalla Prakash, CIE
• 11 a.m – 12 p.m │ Sustainable ML Architecture │ Ms Akhila Matathammal
• 12 p.m. – 1 p.m.│ Valediction

Legend: L – Lecture, T – Tutorial, D – Discussion

Resource Persons

NB: Admission is restricted to an invited set of participants.

Coordinators (logistics support)

Event Photos:

Hotels nearby IIIT Hyderabad:

S. No.	Name of the Hotel	Address	Distance to (IIIT-H)
01	Aditya INN	Plot No.4, IDBI Bank, Sri Harsha Apartments, Beside JN, Vinayak Nagar, Indira Nagar, Gachibowli, Telangana 500032 Contact Persons: Nagendar: 83740 52233, Sridhar: 93486 38685 Email: adityainn@hotmail.com Phone: 040-42000518, 40138108	900 M
02	Ellaa	Hill Ridge Springs, Gachibowli, Hyderabad, Telangana – 500032	1.9 KM
03	Red Fox	Plot No. 2, Survey No. 64, Hitech City, Madhapur, Hyderabad, Telangana 500081	4 KM
04	Radisson	Gachibowli Miyapur Rd, Hyderabad, Telangana 500032	4 KM
05	Treebo Tara Residency	Madhapur	4 KM
06	Lemon Tree	Survey No. 115/1, Financial District, Nanakram Guda, Gachibowli, Hyderabad, Telangana 500032	4.5 KM
07	Westin	Building No: 15, Raheja IT Park, Hitec City, Madhapur, Hyderabad, Telangana 500081	6 KM
08	Sheraton	115/1 Nanakramguda, Financial District, Gachibowli, Hyderabad, Telangana, 500032	6.7 KM
09	Trident	HITEC City, Near Cyber Towers, Madhapur, Hyderabad, Telangana, 500081	10 KM
10	Lime Boutique	4-50/47, Jayabheri Enclave, Gachibowli, Hyderabad, India	2.6 KM

For reservations, participants are advised to please contact the hotel directly.

Participants List

Sl No.	Lab	Name	Designation
1	ACEM	Vinay Siram	Sc C
2	ARDE	Sh. Chandra Prakash Singh	Sc E
3	ASL	Sh. Komati Raghunath Rao	Sc F
4	ASL	Sh. Venkat M	Sc F
5	CEMILAC	Ms. Swapna CS	Sc E
6	DEAL	Sh. Sandeep Kishore	Sc E
7	DLRL	Sh. Sanjay Pandav	Sc G
8	DLRL	Sh. Omkar Kothekar	Sc E
9	DMRL	Smt. Bhongiri Prasuna	Sc B
10	DRDE	Sh. Sunil Agrawal	Sc E
11	DRDL	Sh. Akula Naresh	Sc F
12	DRDL	Sh. Vaddi Chandra Sekhar	Sc C
13	RCI	Sh. Tulip Kumar Toppo	Sc F
14	DYSL-SM	Dr. Rashmi	Sc E
15	IRDE	Sh. Manoj Purohit	Sc F
16	ISSA	Dr. V G Patil	Sc G
17	ISSA	Ms. Chitra Kushwaha	Sc B
18	LRDE	Sh. Jang Bahadur Singh	Sc F
19	MSC	Sh. Sopan B Gosavi	Sc B
20	IRDE	Sh. Nitin Ranjan	Sc E
21	LRDE	Sh. Vishwakarma Rahul Lalmani	Sc E
22	NSTL	Sh. CH. V Satya Srinivas	Sc G
23	PXE	Sh. Suman Mondal	Sc D
24	SAG	Ms Akansha Yadav	Sc C
25	RCI	Smt. K Sravani	Sc D
26	SAG	Shantanu	Sc E
27	SAG	Kuldeep Goyal	Sc E
28	SPIC	Sh. Manak Babra	Sc D
29	SSPL	Sh. Pankaj Kumar	Sc E
30	TBRL	Sc. Sh. Priyanshu Singh	Sc B
31	MSC	Sh. Aditya Kalsha	Sc B

DRDO Labs and Establishments Participated:

•Advanced Centre for Energetic Materials
•Armament Research & Development Establishment
•Advanced Systems Laboratory
•Centre for Military Airworthiness and Certification
•Defence Electronics Applications Laboratory
•Defence Electronics Research Laboratory
•Defence Metallurgical Research Laboratory
•Defence Research & Development Establishment
•Defence Research & Development Laboratory
•DRDO Young Scientists Laboratory for Smart Materials
•Instruments Research & Development Establishment
•Institute for Systems Studies & Analyses
•Electronics & Radar Development Establishment
•Mobile System Complex
•Naval Science & Technological Laboratory
•Proof and Experimental Establishment
•Research Centre Imarat
•Scientific Analysis Group
•System Planning & Implementation Centre
•Solid State Physics Laboratory
•Terminal Ballistics Research Laboratory

Attendance Analysis

Program Certificate

Analytics on questions involving levels of satisfaction – Feedback

1. Levels of Satisfaction with the Advanced Training Program on AIML for DRDO (December 09–20, 2024)

Analysis of Responses:

• Positive Feedback:
  • The program provided participants with exposure to advanced AI/ML concepts and applications, particularly those relevant to defence.
  • Many participants appreciated the opportunity to interact with experts and gain insights into state-of-the-art AI models and technologies.
• Suggestions for Improvement:
  • Some participants found the course insufficiently tailored to DRDO’s specific needs and practical applications.
  • Requests for more niche and targeted topics, along with basic courses for beginners, were common.
  • Suggestions included periodic training sessions and a mix of executive and certification programs to address various levels of expertise.
• Overall Satisfaction:
  • Approximately 85% of participants expressed satisfaction, while 15% suggested scope for improvement, specifically in aligning content with DRDO-specific challenges.

2. How Do You Envisage DRDO as an Organization to Benefit from Programs Like This One in the Next Ten Years? What Should DRDO Be Doing to Achieve Such Goals?

Analysis of Responses:

Long-Term Benefits for DRDO:

• Exposure and Applications: Programs like this provide participants exposure to advanced AI/ML fields, enabling the use of knowledge for defense applications and ensuring DRDO remains globally competitive.
• Technological Advancements: State-of-the-art models from IIIT Hyderabad and collaboration with academia can drive innovation and ensure DRDO keeps pace with global developments.
• Infrastructure and Collaboration: Establishing AI technology centers in each lab, along with a centralized AI nodal lab for infrastructure and research support, will facilitate problem-solving and knowledge-sharing across DRDO.

2. Suggestions for Achieving Goals:

• Upskilling and Training:
  • Frequently conduct programs ranging from basic to advanced levels, covering broader participation.
  • Provide periodic upskilling to ensure scientists are aware of emerging technologies.
  • Develop certification programs for junior scientists and executive courses for senior scientists.
• Infrastructure Development: Establish center-level hardware and supporting infrastructure to train and apply AI/ML technologies.
• Open-Minded Approach: Promote open-mindedness to adopt and integrate new technologies effectively.
• Knowledge Sharing: Encourage collaborative efforts across labs to share advancements and avoid siloed work.

3. Benefits of Continued Training Programs:

• Enhances understanding of the latest AI technologies.
• Drives productivity and data-driven decision-making.
• Develops better products and solutions for end users through technology awareness.
• Supports automation while ensuring data privacy and reduced human intervention.

4. Opportunities for Improvement:

• Include tailored courses on niche and targeted topics for advanced learners.
• Basic AI/ML courses should be accessible to a larger pool of scientists to build foundational knowledge.
• Ensure that future courses meet participant expectations and provide high-quality learning experiences.

5. Collaborative Vision:

• Collaborate with prestigious institutes to make periodic upskilling a continuous process.
• Build frameworks and guidelines for applying AI/ML in mission-critical systems while maintaining confidentiality.
• Leverage collective knowledge and state-of-the-art technologies for defense advancements.

3. Continuous Knowledge Updation: How Can Your Centre/Cluster Help You Here?

Analysis of Responses:

1. Importance of Knowledge Updation:

• Continuous knowledge and skillset updation is vital to remain at the forefront of Machine Learning advancements.
• It ensures scientists can implement modern AI/ML techniques effectively in their projects and research.

2. Suggestions for Support from Centers/Clusters:

• Organize Regular Training and Workshops:
  • Conduct in-house courses and hands-on training on evolving AI architectures, including self-learning and generative AI.
  • Periodically nominate scientists for training programs at prestigious institutes like IIIT Hyderabad to stay updated.
• Provide Research Facilities:
  • Equip labs with necessary hardware, software, and high-speed internet to facilitate research and learning.
  • Relax procurement policies for faster access to required hardware and resources.
• Promote Collaboration:
  • Foster collaboration between labs and academic institutions to share knowledge, resources, and applications.
  • Build platforms for inter-lab interactions and engagements with academia for joint projects.
• Create a Supportive Environment:
  • Allocate dedicated research time for scientists to explore and apply AI/ML in their domains.
  • Encourage participation in conferences and workshops to enhance learning and networking opportunities.

3. Key Suggestions for Implementation:

• Disseminate application areas of AI/ML and create awareness about their practical implications.
• Provide tools and infrastructure necessary for independent research and experimentation.
• Align learning programs with industry trends and real-world applications to ensure relevance.

4. Benefits Highlighted:

• Application of new knowledge in experimental projects can lead to innovation in defence technologies.
• Shared experiences and collaborative efforts can lead to breakthroughs in solving real-world problems.
• Regular interaction with academia and participation in skill-building initiatives can enhance productivity and innovation.

5. Final Recommendation:

• Centers and clusters should focus not only on knowledge updation but also on skillset enhancement by providing cutting-edge tools, fostering collaboration, and ensuring alignment with emerging technologies and trends.

4. Participation in Conferences and Symposiums at IIIT Hyderabad

1. Strong Agreement on Participation:

• The majority of respondents believe participation in short-duration programs, conferences, and symposiums at IIIT Hyderabad is highly beneficial for both individuals and the organization.
• There is unanimous support for these programs as opportunities to stay updated, enhance knowledge, and align with advancements in the fast-evolving AI/ML field.

2. Benefits Highlighted for DRDO:

• Knowledge Enhancement:
  • Participants gain insights into the latest research, trends, and technologies in AI/ML.
  • It boosts confidence and understanding, enabling better implementation of advanced concepts in lab projects.
• Networking Opportunities:
  • Facilitates interaction with researchers and experts, promoting collaboration and exchange of ideas.
• Time-Saving for Research:
  • Access to focused knowledge and practical takeaways from these events helps save time when exploring new technologies for specific tasks.
• Lab Productivity:
  • Applying learnings from these events can improve lab projects’ performance and contribute to task efficiency.

3. Suggestions for Maximizing Impact:

• Online Accessibility:
  • Organizing conferences and symposiums in an online format would allow DRDO employees to participate without disrupting their work schedules.
• Advance Notifications:
  • Providing early information and comprehensive details about such events would help participants plan effectively and ensure high attendance.
• Relevance to DRDO Work:
  • Focus on conferences that align with DRDO’s ongoing projects and domains of interest to ensure maximum applicability of learnings.

4. Areas of Focus for Future Events:

• Topics such as Transformers, Large Language Models (LLMs), and Prompt Engineering are of significant interest and relevance for participants.
• Narrow, domain-specific topics within AI/ML that address emerging challenges and trends.

5. Final Recommendation:

• IIIT Hyderabad should continue organizing such programs and ensure broader dissemination of event details, including online options.
• Regular participation in these events will help DRDO stay aligned with advancements, foster collaboration, and enhance organizational capabilities in AI/ML.

Summary of Feedback and Insights:

The Advanced Training Program on AI/ML for DRDO at IIIT Hyderabad has been highly valued for its ability to expose participants to cutting-edge AI/ML concepts, foster collaboration, and align DRDO with global advancements in technology. While 85% of participants expressed satisfaction, there is a recognized need to tailor future programs more closely to DRDO-specific challenges, with a mix of beginner, advanced, and niche topics.

Key recommendations include:

1. Continuous Upskilling: Regular training, workshops, and certification courses at all expertise levels, alongside participation in conferences and symposiums, are essential for staying updated in the rapidly evolving AI/ML domain.
2. Enhanced Infrastructure at DRDO: Establish state-of-the-art labs within DRDO, equipped with advanced hardware, software, and high-speed internet, to support research, experimentation, and the effective application of AI/ML technologies.
3. Collaboration: Strengthen ties with academic institutions like IIIT Hyderabad and promote inter-lab interactions to share advancements and work on real-life problems.
4. Practical Implementation: Ensure takeaways from training programs are directly applicable to DRDO projects, fostering innovation and boosting lab productivity.

By adopting these strategies, DRDO can stay at the forefront of AI/ML advancements, ensuring long-term technological competitiveness and operational efficiency.