Student Activities

I am delighted to acknowledge the tremendous efforts and dedication of the students who participated in my courses. Their sincerity, hard work, and contributions have significantly enriched the learning experience for everyone involved. As part of these course, students have been engaged in "Course Project" component, which allowed them to apply theoretical knowledge to practical problems. Students either worked individually or collaborated in groups to explore diverse topics, demonstrating creativity, critical thinking, and technical skills. Their projects covered a wide range of applications, including but not limited to machine learning, natural language processing, computer vision, and robotics. I am incredibly proud of their achievements and the innovative solutions they developed. Below are the details of group and their projects are provided for further reference.

DA 675 Fuzzy Systems and Applications (JUL - NOV 2024)

Uploading soon...

DA 241M Mathematical Foundations of Data Science (JUL - NOV 2024)

Uploading soon...

DA 221M Artificial Intelligence (JAN - MAY 2024)

Group Name: TAJ

Group Members:
Tanvi Doshi (IITG Roll No.: 220101102, Email: t.doshi@iitg.ac.in)
Adarsh Gupta (IITG Roll No.: 220101003, Email: adarsh.gupta@iitg.ac.in)
Japleen Kaur (IITG Roll No.: 220106035, Email: japleen@iitg.ac.in)

Project Description: This project aims to carry out binary and multi-class classification of Knee Osteo-Arthritis severity. It utilizes a stacked ensemble of deep learning models, leveraging transfer learning on a diverse range of pre-trained state-of-the-art architectures including MobileNetV2, YOLOv8, DenseNet, EfficientNet, CvT, and ResNet50. The models are trained on the Osteoarthritis Initiative (OAI) dataset and the classification is based on the Kellgren-Lawrence scale.

GitHub Code Link: Click here

Group Name: Rythm

Group Members:
D Hayagrivan (IITG Roll No.: 220101032, Email: d.hayagrivan@iitg.ac.in)
Kushal Gupta(IITG Roll No.: 220123082, Email: g.kushal@iitg.ac.in)

Project Description: This project aims to produce music by predicting upcoming notes using an LSTM model. The model is initialized with the first few notes and then generates music based on the training set. It is trained on a pre-labeled dataset called "MusicNet." By using patterns learned from MusicNet, the model creates sequences of notes. This approach enables the generation of music that follows the style of the training data. The project showcases the potential of artificial intelligence in music composition.

GitHub Code Link: Click here

Group Name: SVP

Group Members:
Shreyas Manish Saxena (IITG Roll No.: 220108073, Email: s.shreyas@iitg.ac.in)
Sapaharam Vishnu Singh (IITG Roll No.: 220108053, Email: s.sapaharam@iitg.ac.in)
Priyanshu Maurya (IITG Roll No.: 220108046, Email: priyanshu.maurya@iitg.ac.in)

Project Description: A movie recommender system has been developed using two approaches. The first approach is a weighted hybrid technique, where a formula was applied to calculate weights for movies based on their data. Higher weights result in higher recommendation ranks. In the second approach, content-based filtering was used, where relevant features were extracted from movie summaries and compared between pairs of movies. This comparison produced a similarity score, with higher scores indicating closer relationships between movies. For any given movie, a list of closely related movies can be recommended based on these scores. These methods enhance the accuracy and relevance of recommendations.

GitHub Code Link: Click here

Group Name: 7ARMS

Group Members:
Arush Shaleen Mathur (IITG Roll No.: 220101017, Email: m.arush@iitg.ac.in)
Ayush Kumar (IITG Roll No.: 220101021, Email: ayush.cse22@iitg.ac.in)
Ayush Savar (IITG Roll No.: 220101022, Email: a.savar@iitg.ac.in)
Dev Shah (IITG Roll No.: 220101035, Email: dev.shah@iitg.ac.in)

Project Description: This project aims to develop a method for estimating the number of people in images taken from arbitrary perspectives. Estimating crowd density is useful for applications such as crowd control, public safety, infrastructure planning, and resource deployment. The neural network architecture is split into three parallel CNNs, each with filters of different sizes to capture features at various scales. Max pooling is applied to each 2×2 region, and ReLU activation functions are utilized to enhance CNN performance. This approach ensures accurate identification and counting of heads in images with varying perspectives and scales.

GitHub Code Link: Click here

Group Name: A.C.R.S

Group Members:
Sharanya Adane (IITG Roll No.: 220123080, Email: sharanya@iitg.ac.in)
Shubham (IITG Roll No.: 220103103, Email: shubham.mech22@iitg.ac.in)
Simon Lalremsiama Shangpliang (IITG Roll No.: 220101095, Email: s.shangpliang@iitg.ac.in)
Aryan Adlakha (IITG Roll No.: 220101123, Email: a.adlakha@iitg.ac.in)

Project Description: In response to the prevalent challenge of cyberbullying, this project is dedicated to developing a solution that effectively identifies and addresses instances of online harassment. Focusing on Natural Language Processing (NLP), the approach centers on analyzing tweets to detect signs of cyberbullying. Central to the methodology is the utilization of BERT (Bidirectional Encoder Representations from Transformers), a cutting-edge pre-trained model renowned for its prowess in understanding and processing natural language data. Trained on relevant datasets, the implementation of BERT aims to discern patterns and characteristics indicative of cyberbullying within tweets. This project empowers users and platform administrators to take proactive measures against online harassment.

GitHub Code Link: Click here

Group Name: SYSS

Group Members:
Srinjoy Som (IITG Roll No.: 220123074, Email: s.srinjoy@iitg.ac.in)
Surankan De (IITG Roll No.: 220101121, Email: d.surankan@iitg.ac.in)
Yash Jain (IITG Roll No.: 220101115, Email: j.yash@iitg.ac.in)
Shubham Kumar Jha (IITG Roll No.: 220123081, Email: shubham.jha@iitg.ac.in)

Project Description: This project aims to develop a movie recommendation system that suggests movies similar to those a user searches for. For generic movie recommendations, users will receive suggestions based on movie ratings and popularity. When a specific movie is inputted, the suggestion engine will recommend movies similar in terms of overview, as well as provide separate recommendations based on cast, crew, and other attributes. In essence, this hybrid recommendation system incorporates demographic, content-based, and collaborative filtering techniques to deliver personalized movie suggestions.

GitHub Code Link: Click here

Group Name: A.M.A.N

Group Members:
Manas Jhawar (IITG Roll No.: 220123032, Email: j.manas@iitg.ac.in)
Aniket Gupta (IITG Roll No.: 220123008, Email: g.aniket@iitg.ac.in)
Arush Jain (IITG Roll No.: 220101016, Email: j.arush@iitg.ac.in)
Nilesh Singla (IITG Roll No.: 220123078, Email: s.nilesh@iitg.ac.in)

Project Description: This project develops a music genre classifier using Convolutional Neural Networks (CNN) and Convolutional Recurrent Neural Networks (CRNN) with Mel-frequency cepstral coefficients (MFCCs) from mel spectrograms. The GTZAN dataset, with 1000 songs across 10 genres, is used. Audio data is preprocessed to generate mel spectrograms, and MFCC features are extracted. A CNN processes MFCC features directly, while a CRNN captures spatial and temporal information. The models are trained and evaluated using the GTZAN dataset, with accuracy as a key performance metric. Hyperparameter tuning and optimization enhance classification accuracy.

GitHub Code Link: Click here

Group Name: idunit

Group Members:
Rathod Sujal Jignesh (IITG Roll No.: 220102077, Email: r.jignesh@iitg.ac.in)
Patel Dhyey Sanjaykumar (IITG Roll No.: 220101077, Email: pd.sanjaykumar@iitg.ac.in)
Sharvil Viral Patel (IITG Roll No.: 220101091, Email: p.sharvil@iitg.ac.in)
Patel Czileen Natvarbhai (IITG Roll No.: 220101075, Email: p.natvarbhai@iitg.ac.in)

Project Description: This project utilizes an LSTM model for predicting stock prices based on historical trends. Stocks from motor companies were selected and imported using the yfinance library to ensure uniformity. The data was meticulously preprocessed by removing outliers, eliminating null values, and normalizing it to enhance model performance. These preprocessing steps facilitate achieving optimal weights in a small number of epochs. The final outcomes were plotted to evaluate and visualize the model's effectiveness in predicting stock prices.

GitHub Code Link: Click here

Group Name: phytoez

Group Members:
Aaditya Jalan (IITG Roll No.: 220123001, Email: j.aaditya@iitg.ac.in)
Abhishek Kumar (IITG Roll No.: 220101002, Email: abhishek.cse22@iitg.ac.in)
Gadhikar Avanish Harshal (IITG Roll No.: 220123075, Email: h.gadhikar@iitg.ac.in)
Sanchay Baranwal (IITG Roll No.: 220101122, Email: b.sanchay@iitg.ac.in)

Project Description: This project aims to develop an end-user application for classifying and detecting plant diseases through image classification. The application will categorize images of plants into healthy and unhealthy categories. Unhealthy plants will be further classified into various specific diseases, enabling accurate disease detection and diagnosis.

GitHub Code Link: Click here

Group Name: Team Summerizers

Group Members:
Chinam Suraj Patra (IITG Roll No.: 220102107, Email: c.patra@iitg.ac.in)
Tanay Goenka (IITG Roll No.: 220101098, Email: t.goenka@iitg.ac.in)
Priyanshu Pratyay (IITG Roll No.: 220101125, Email: p.pratyay@iitg.ac.in)
Subhrajyoti Kunda Roy (IITG Roll No.: 220123064, Email: r.subhrajyoti@iitg.ac.in)

Project Description: This project involved developing an automated research paper categorization system using Deep Learning and NLP algorithms. The system analyzes content and structure to accurately categorize papers into predefined topics, aiding in organizing academic libraries, streamlining journal editorial processes, and enhancing literature reviews and conference management. Data preprocessing, class weighting, and resampling techniques were used to address imbalances. A Dense Neural Network with Swish activation functions and Binary Cross-Entropy loss was constructed, optimized with the Adam optimizer, and evaluated using the F1 score. The system provides significant benefits for academia, industry, and policy analysis.

GitHub Code Link: Click here

Group Name: AK

Group Members:
Kovid Juneja (IITG Roll No.: 220101059, Email: j.kovid@iitg.ac.in)
Ankit Varshney (IITG Roll No.: 210121069, Email: ankit.varshney@iitg.ac.in)

Project Description: This project aims to improve systolic blood pressure (SBP) prediction accuracy using a hybrid model combining Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) architectures. The LSTM component captures temporal dependencies, modeling the sequential nature of blood pressure measurements. The CNN component extracts relevant spatial features from the data. By integrating these architectures, the model aims to enhance SBP predictive capabilities. This approach has the potential to improve the diagnosis and treatment of hypertension, contributing to better healthcare outcomes and patient well-being.

GitHub Code Link: Click here

Group Name: Stochastic Strikers

Group Members:
Ghosh Kushanava Amitava (IITG Roll No.: 220123083, Email: g.amitava@iitg.ac.in)
Aviral Srivastava (IITG Roll No.: 220102109, Email: aviral.s@iitg.ac.in)
Shashwat Shankar (IITG Roll No.: 220101092, Email: s.shankar@iitg.ac.in)
Savani Shivam Bhagwanbhai (IITG Roll No.: 220101090, Email: b.savani@iitg.ac.in)

Project Description: Two well-known Naive Bayes models were implemented for Email Spam Classification: Bernoulli's Event Model (BEM) and Multinomial Event Model (MEM). The performance of both models was compared, and it was concluded that MEM is more efficient and powerful for email spam classification than BEM. The project demonstrates the application of these models and their effectiveness in distinguishing spam emails from legitimate ones.

GitHub Code Link: Click here

Group Name: Huh?

Group Members:
Krish Mangal (IITG Roll No.: 220102051, Email: m.krish@iitg.ac.in)
Prachi Bindal (IITG Roll No.: 220102071, Email: p.bindal@iitg.ac.in)
Kushagra Singh Sisodia (IITG Roll No.: 220102052, Email: s.kushagra@iitg.ac.in)
Tanu Siwach (IITG Roll No.: 220108059, Email: t.siwach@iitg.ac.in)

Project Description: This project focuses on accurately identifying the boundaries between exons and introns within DNA sequences, which is crucial for gene structure understanding, protein prediction, and genetic disorder diagnosis. DNA sequences, represented by 180 binary variables, are labeled as Exon-Intron (EI), Intron-Exon (IE), or Neither. Ensemble learning techniques, including Random Forest, Gradient Boosting, and Stacking, are utilized for model training and evaluation. The performance of these models is compared using accuracy, precision, recall, and F1-score metrics. This project aims to advance genetic research by improving the accuracy of exon-intron boundary identification.

GitHub Code Link: Click here

DA 213 Python Programming Laboratory (JAN - MAY 2024)

Group Name: Zero Proxy

Group Members:
Shivansh Pal (IITG Roll No.: 220150022, Email: p.shivansh@iitg.ac.in)
Karnati Ravi Teja (IITG Roll No.: 220150006, Email: karnati@iitg.ac.in)
Mohit Yadav (IITG Roll No.: 220150008, Email: mohit.dsai22@iitg.ac.in)

Project Description: An attendance record system using face recognition in Python has been developed. Initially, images of all students with their roll numbers were taken to serve as labels. Faces were then extracted, aligned, and measured using OpenCV functions. During operation, the system camera captures a student's face, processes it, and calculates the Euclidean distance to the trained images. Attendance is marked if the distance is within the threshold.

GitHub Code Link: Click here