Farewell of B-Tech
Artpark IISc Bangalore Summit
Robotics Challenge 2022
Microsoft Workshop at NIT Srinagar
Vision Group IIT DELHI

Tajamul Ashraf


I am currently pursuing my Masters in Research from School of Information Technology at Indian Institute of Technology Delhi under the guidance of Prof. Chetan Arora. I completed my bachelors from the Department of Information Technology at National Institute of Technology, Srinagar.

My research interests lie at the intersection of Computer Vision, Robotics and Machine Learning.

If you cannot do great things, do small things in a great way.

Napolean Hill

About memy stats


Tajamul Ashraf received his B. Tech. degree in Information Technology from the National Institute of Technology Srinagar and is currently an MS(R) student at Indian Institute of Technology Delhi. He completed his schooling from Tyndale Biscoe School.

He has previously worked at Leena AI as a software engineer. He is also founder of an organization RALITH MILITH which is an anti drug society based in Kashmir. Recently he co-founded an podcast channel ATTalks which throws light on various technological interventions surrounding us.

For more information, see my Resume.









Languages I Know













My Timeline


Software Developer - Leena AI

The responsibilities included executing full lifecycle softwares, architecting the solutions to meet client requirements, working in cross -functional teams(Product, Engineering and ML team).

2022 - present


Currently coursework going on.
Teaching Assistant (TA) for
COL106- Data Structures and Algorithms,
COl100- Introduction to Computer Science,
COL780- Computer Vision.

Jan 2022 - May 2022

Computer Vision Reseacher - Artpark IISc

Mentor: Prof. Raghu Krishnapuram

The challenge requires a robot to demonstrate janitorial tasks that would be typically performed in a washroom.

2018 - 2022


Mentor: Dr. Janibul Bashir

B-tech Thesis: A Novel Robust Deep Learning Model to find the effect of variations of multiple coherent climatic factors on Climate Change.

Nov 2019 - Mar 2020

Research Intern - IIT KHARAGPUR

Mentor: Prof. Pabitra Mitra

Our Goal was to use computer vision to aid agricultural robotics. One problem was identifcation of fruits in a tree. Devise an algorithm for counting fruits. Implement the algorithm and run on data set. Devise an algorithm to identify fresh and rotten fruit..

2016 - 2018

Tyndale Biscoe School - Higher Secondary

Term Project: IoT based flood monitoring and alerting system in Jhelum river.

Class XII with 91.8%

Class X with 92.6%

Featured In

Meet a recent Microsoft Learn Student Ambassador graduate.

This is the next segment of our blog series highlighting Tajamul Ashraf who achieved the Gold milestone, the highest status level attainable, and recently graduated from university.- MICROSOFT

NIT Srinagar students win India’s largest National Entrepreneurship Challenge

NEC was a 6 month-long and a highly competitive contest which was led by the team leader Mr Tajamul Ashraf- GREATER KASHMIR

Workshop on entrepreneurship, research and Innovation held at NIT Srinagar

The workshop was successfully coordinated by Tajamul Ashraf, Sanna Showkat and other IIEDC student team members- KASHMIR READER

Workshop on Emerging Technologies held at NIT Srinagar

While organizing team consisted of Basar Qari, Sanna Showkat, Manik Showkat, Fahad Syed, Ubaid bin Gulzar and Mir Faizan. While it was headed by the Microsoft Student Learn Ambassador for NIT Srinagar, Tajamul Ashraf- SCOOP NEWS

HULT PRIZE: A futuristic approach to entrepreneurship

You have to see failure as the beginning and the middle, but never entertain it as an end. - KASHMIR MONITOR

IEEE organizes career counseling session on GRE preparation at NIT Srinagar

In his message, Director NIT Srinagar, Prof (Dr.) Rakesh Sehgal appreciated the organizers for initiating such ideas for the guidance of aspirants- RISING KASHMIR

ResearchMy Work

My research interests lie at the intersection of Computer Vision, Robotics and Machine Learning.



  1. Tajamul Ashraf, Fuzayil-bin-Afzal Mir, Iqra Altaf Gillani, A way to epitomize Transformer based Focal Modulation using Federated Learning. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2024).


  2. 2023

  3. Tajamul Ashraf, Balaji Prabu BV, and Omkar Subbaram Jois Narasipura PoseWatch: Advancing Real Time Human Pose Tracking and Juxtaposition with Deep Learning, International Conference on Computer Vision and Image Processing (CVIP 2023).


  4. Tajamul Ashraf, Janibul Bashir Climate Change Parameter Model (CCPM): A Robust Deep Learning Approach for Weather Forecasting and Climate Change Analysis, Neural Computing and Applications 2023.


  5. Tajamul Ashraf, Janibul Bashir Climate Change Parameter Dataset (CCPD): A Benchmark Dataset for Climate Change parameters in Jammu and Kashmir,, In proceedings of ICDSA 2023.


  6. 2022

  7. Tajamul Ashraf, Naiyer Abbas, Mohammad Haseeb, Nadeem Yousuf, Janibul Bashir, 2022. An Integral Computer Vision System for Apple Detection, Classification, and Semantic Segmentation. In Proceedings of the 15th International Conference on Machine Vision (ICMV 2022).

    [Paper] [Code]

  8. Tajamul Ashraf, Fuzayil-bin-Afzal Mir, 2023. A proposed scale to measure the effect of team building on Engineering students. International Conference on Recent Advances on Business, Management and Data Analystics 2023.



A Novel Robust Deep Learning Model to find the effect of variations in multiple coherent climatic factors on Climate Change.

Mentor: Dr. Janibul Bashir

One of the severe and pressing problems that the human race is currently facing is the climate change. It is adversely affecting the entire globe as is evident with the rising ocean levels, accelerated melting of glaciers, frequent storms, and many more Urbanization and infrastructure development are the main reasons that the environment is getting polluted, resulting in a frequent change in weather patterns that have led to flash floods, landslides, and avalanches in Jammu and Kashmir in the past few years. In this paper, we have tried to address this problem by developing a robust climate change model that can capture the different parameters that are responsible for accelerating the climate change. We have intelligently processed the data related to pir panjal range collected over a long period of time. Subsequently, we developed a sequential deep learning-based Climate Change Parameter model to understand the correlations between the parameters and their contribution to climate change. We proposed a Stacked (Long Short-Term Memory) LSTM-based learning model called Climate Change Parameter Model (CCPM) to estimate the variables which are directly or indirectly affecting climate change. This helps us in formulating further superlative policies, univariate and multivariate forecasting, and efficient strategies to cope with the problem. We evaluated our model on the dataset and found the model loss within 2%.

B-Tech Project

A Integral Computer Vision System Using Recoginization, Classification and Semantic Segmentation to Aid Agricultural Farm Robotics.

Mentor: Prof. Pabitra Mitra

The area of Computer Vision has gone through exponential growth and advancement over the past decade. It is mainly due to the introduction of effective deep learning methodologies and the availability of massive data. This has resulted in the incorporation of intelligent computer vision schemes to automate the different number of tasks. In this paper, we have worked on similar lines. We have proposed an integrated system for the development of robotic arms, considering the current situation in fruit identification, classification, counting, and generating their masks through semantic segmentation. The current method of manually doing these processes is time-consuming and is not feasible for large fields. As a result, multiple works have been proposed to automate these tasks to reduce the overall overhead. However, there is a lack of an integrated system that can automate all these processes together. As a result, we are proposing one such system based on different machine learning techniques. For each process, we propose to use the most effective learning technique with computer vision capability. Thus, proposing an integrated intelligent end-to-end computer vision-based system to detect, classify, count, and identify the apples. In this system, we modified the YOLOV3 algorithm to effectively detect and count the apples. The proposed scheme works even under variable lighting conditions. The system was trained and tested using a standard benchmark i.e., MinneApple. Experimental results show an average accuracy of 91%


Vision problems in ARTPARK’s Robotics Challenge 2021-22 (Robojanitor)

Mentor: Prof. Raghu Krishnapuram

The ARTPARK Robotic Challenge 2021-22 required a robot to demonstrate janitorial tasks that would be typically performed in a washroom. In this challenge, the tasks were limited to clearing any rubbish that may be on the floor followed by cleaning the washbasin and the washbasin counter using a sanitizing liquid. The rubbish consisted of items such as tissue paper and small paper cups and cans. The robot had to scout out the restroom area so that it can assess the environment and create a representation of it for planning and navigation purposes. The tasks which were to be completed were Detecting the trash from the floor-The robot needed to detect trash on the floor The trash consisted of items such as crumpled tissue paper, paper cups, and cans. Pose estimation of countertop -The robot needs to dry the wet countertop and clean it Cleaning the floor- The robot needed to spray sanitizing liquid on the floor on dirty spots and then clean them up. Finally, the robot had to mop the floor of the restroom. Figure 1 shows the setup of the washroom where different tasks were performed.

ARTPARK Robotics Challenge

Human Pose Tracking and Juxtaposition(HPTJ) Model: Pose Comparisom and real time evaluation and metric.

Mentor: Prof. S N Omkar, Dr. Balaji Prabhu

Pose estimation is a computer vision technique that predicts and extracts the Key-point location of a person or object to track the actions made by the person while performinig any activity to draw the insights. This paper presents a novel and state of the art Human Pose Tracking and Juxtaposition(HPTJ) Model It is the first of its kind where in real time we propose a Instructor Based Model, which gives proper instructions to the user and provides him/her with a feedback of the whole. The model was developed with an aim to aid the current Instructors and Free lancers in their tracking methods and also for the students to improve their progress. This paper lays the foundation for building a system that can recognise human pose on pre- recorded videos as well as in real-time. It also has inbuilt features of giving instructions to correct the posture of users. The various applications of the HPTJ model is in Activity Identify applicable funding agency here. If none, delete this. Recognition, where tracking the variations in the pose of a person over a period of time can also be used for activity, gesture and gait recognition. Model can autonomously teach proper work out regimes, sport techniques and dance activities. Also it can be used to understand full-body sign language, motion Capture and augmented reality and Training Robots.

IISc Internship

Develop a network simulator implementing entire protocol stack.

Mentor: Dr. Iqra Altaf Gillani

A simulated model which demonstrates the working of the TCP/IP Networking Model. This project shows proper working of the five layers: 1.Physical Layer: The physical layer is responsible for sending and receiving packets from one physical interface to others 2.Data Link Layer: The data link layer controls the flow of reception, delimitation and transmission of frames and establishes a communication protocol (Media Access Control) between directly connected systems. 3.Network layer: Handles the routing and sending of data between different networks. The most important protocols at this layer are IP and ICMP.We have implemented the Routing Information Protocol-(RIP) in the project for shortest path. 4.Transport layer: Provides the means for transmitting data between the two connected parties, as well as controlling the quality of service. The main protocols used here are TCP and UDP.We have set a probability for packet drop in UDP to show that this protocol is unreliable. TCP is a reliable protocol where no packet drop is shown.A server class is made to enable the 3 services we have included HTTP, SSH, SMTP. 5.Application Layer: The application layer is the highest abstraction layer of the TCP/IP model that provides the interfaces and protocols needed by the users. It combines the functionalities of the session layer, the presentation layer and the application layer of the OSI model.We have included two application layer services DHCP and HTTP

Networking Project

Robot localization and Mapping using Particle Swarm optimization (PSO)

Mentor: Dr. Iqra Altaf Gillani

The goal of Wireless Network Systems (WSN) Localisation is to find the position of the target node using the priori knowledge of the absolute position of the n anchor nodes and the distance between anchor and target nodes. The distance (range) can be determined using received signal strength (RSS), Angle of arrival (AOA) or any of the time based techniques like Time of arrival (TOA), Time difference of arrival (TDOA) or Round trip time (RTT). The Position estimation of the coordinates of the target nodes can be formulated as an optimization problem, involving minimization the of an objective function representing locating the error in the target nodes. The sum of squared range errors between the target nodes and neighboring anchor nodes can be considered as the objective function for this problem. The coefficients have a critical role in the convergence characteristics of PSO. The coefficient w should be neither too small, which results in early convergence, nor too large, which on contrary slows down the convergence process. A value of w = 0.7 and c1 = c2 = 1.494 were recommended for faster convergence by [1-2] after experimental tests. Four Target nodes were fixed at (0,100), (0,-100), (100,0) and (-100,0). Anchor nodes are generated randomly and error is calculated as distance between between actual position and estimated position.

Natural Algorithms project

Encryption and Decryption between users in Morse using a Flashlight.

Mentor: Exposys Data Labs

Morse code is a process of transmitting text information as a series of on-off tones and lights or clicks. If they use a tapping device recipient can understand the message without additional decoding equipment. Morse code is represented by the form of dits and dahs. Here dits refer to dots and dahs refer to dash. Morse code used to transmit only numerals at first. After that, Alfred Vail included letters and characters. Morse code can be transmitted by using electric telegraph wire, light, and sound, through a different medium in different ways. Tap code is used by American prisoners. Morse code is used for long-distance communication. International Morse code was devised by European nations in 1851. It is the base for the morse code to transmit or receive. Morse code is a character encoding and decoding scheme. This is the implementation of MORSE CODE translator using python. This system was designed to transmit message securely to long distance. The design makes use of a programming language python. This design can be used in different areas like long communication, military, external affairs etc. By using this system there is no need of tapping device for transmission .so this overcomes the security problems.

Encryption and Decryption between users in Morse using a Flashlight.

Assembler which converts all 37 rv32I instructons in hexadecimal and binary machine-code.

Mentor: Dr. Janibul Bashir

RISC-V pronounced as “RISC-five”, is an open-source standard Instruction Set Architecture (ISA), designed based on Reduced Instruction Set Computer (RISC) principles. With a flexible architecture to build systems ranging from a simple microprocessor to complex multi-core systems, RISC-V caters to any market. The RISC-V ISA provides two specifications, one, the User Level Instructions which guides in developing simple embedded systems and connectivity applications and two, the Privilege Level Instructions which guides in building secure systems, kernel, and protected software stacks. RISC-V currently supports three privilege levels, viz.. Machine/Supervisor/User, with each level having dedicated Control Status Registers (CSRs) for system state observation and manipulation. In addition, RISC-V provides 31 read/write registers. While all can be used as general-purpose registers, they have dedicated functions as well. RISC-V is divided into different categories based on the maximum width of registers the architecture can support, for example, RV32 (RISC-V 32) provides registers whose maximum width is 32-bits and RV64 (RISC-V 64) provides registers whose maximum width is 64-bits. Processors with larger register widths can support instructions and data of smaller widths. So an RV64 platform supports both RV32 and RV64

Assembler which converts all 37 rv32I instructons in hexadecimal and binary machine-code

Extra StuffThought space

What is Islam

Islam is an Arabic word which means "surrender, submission, commitment and peace." Thus, Islam can be defined as a path to attain complete peace through voluntary submission to the divine will.

Islam is not a new religion. It is the same truth that God revealed to all His prophets throughout history. Islam is both a religion and a complete way of life. Muslims follow a religion of peace, mercy and forgiveness that should not be associated with acts of violence against the innocent.

Get inspired by the life of the Prophet Muhammad (PBUH), see Seerah Pearls by Shaykh Abdul Nasir Jangda.

For detailed information, watch this one.

Qur'an 20;46

"Don't be afraid, I am with you all the time, lsitening and seeing."

Qur'an 40;60

"Call upon me; I will respond to you."

Qur'an 55;38

"Then which of your Lord's favours will you both deny."

Qur'an 93;7

"And He found you lost and guided you."



The Meaning of Life Is Surprisingly Simple

You can make your quest for meaning manageable by breaking it down into three bite-size dimensions; by Arthur C. Brooks

Top Papers in Computer Vision

A curated set of papers which have been reviewed for the latest scoop in AI/ML.

The illustrated guide to a Ph.D.

It's hard to describe it in words; by Matt Might

Ralith MilithMy Blogs

Ralith Milith Website [Link]

Exciting things are on the horizon! We are diligently working on new content and initiatives to further our mission in promoting drug abuse awareness and creating a drug-free society. Stay tuned for updates and announcements that will empower you to be part of the positive change. In the meantime, feel free to explore our website for valuable resources and information on our ongoing efforts.

Contact MeContact

Contact me here

Postal Address

- Karan Nagar, Srinagar J&K


- tajamul21.ashraf[at]gmail[dot]com
- tajamul[at]sit[dot]iitd[dot]ac[dot]in


- 190010

Mobile Number

+91 907 007 1007


-English, Urdu, Hindi, Kashmiri