Technical Magazine 2023-24 - Computer Department
Technical Magazine 2023-24 - Computer Department
Vision
To be known for imparting quality education in computer engineering to serve the
changing needs of global industry.
Mission
To create an environment that fosters technical and professional growth to make
graduates globally competent.
To develop industry ready professionals, researchers, and entrepreneurs to solve real
world problems and societal issues.
PEO2: Adapt latest technological skills to face challenges of the modern computing
industry.
PEO3: Work in finance, healthcare, security and banking sectors as a team member or
a leader with ethical values.
PSO2 Professional Skills: Graduates will be able to design and develop software systems
using front end and back end technologies by following standard software engineering
principles.
PSO3 Professional Skills: Graduates will be able to analyse the algorithms and
implement optimized solutions in the domain of machine learning, natural language
processing, security and cloud computing and Internet of Things.
DEPARTMENT OF COMPUTER ENGINEERING
11. Demonstrate knowledge and understanding of the engineering and management principles
and apply these to one’s own work, as a member and leader in a team, to manage projects
and in multidisciplinary environments. [Project management and finance]
12. Recognize the need for, and have the preparation and ability to engage in independent and
life-long learning in the broadest context of technological change. [Life-long learning]
DEPARTMENT OF COMPUTER ENGINEERING
Editorial Team
The opportunity to present the technical magazine for the years 2023–2024 is an honor
for the Department of Computer Engineering. From our students, we had assembled seminar and
project titles. Our objective in sharing this data is to inspire students and teach them about the
ongoing requirements in the IT business.
The department's goal is to increase technical proficiency among the students. Since the
second year, the department has been working on this. Outside of the classroom, the department
organizes numerous expert talks, seminars, field trips, and learning materials for students .There
are planned activities to make students aware of the present demand for the IT industry. The
results of those efforts are shown in their placement and admission to further education.
Coordinator HOD
Mrs.P.S.Jadhav Dr. S.N.Zaware
DEPARTMENT OF COMPUTER ENGINEERING
Index
Sr. No Content Page
No.
I. PLACEMENT OF ACADEMIC YEAR 2023-24 1
1.Data Lumina
1. Samaaro
2. Hope Foundation-Bajaj
150 Tripathi Janhavi Sandeep 3. Noqs Digital Pvt. Ltd
System Architecture:
The system architecture of the web application for depression assessment and support is
designed to provide a comprehensive and user-friendly platform for users to evaluate their
mental health and access relevant resources. At its core, the architecture comprises several
interconnected modules that handle various functionalities. The system begins with a user
registration and authentication module, ensuring secure access to the application. Upon login,
users are directed to the initial assessment module, where they can complete the BDI-II
questionnaire and optionally provide audio or video submissions to express their feelings and
emotions. The data from these assessments are then processed by the depression severity
analysis module, utilizing machine learning techniques to predict the severity level of
depression. Feedback and personalized recommendations are generated based on these
predictions, guiding users toward appropriate support resources. The system also includes a
video analysis module for emotion detection and additional insights from audio transcription.
Furthermore, users can avail additional features such as scheduling online meetings with
psychologists, accessing 24x7 chatbot support, and maintaining personal journals within the
application. The system architecture leverages a combination of frontend technologies like
HTML, CSS, and Flask, backend technologies including Python and MySQL, machine learning
models for analysis, AI chatbots for immediate assistance, and integration with external
services such as Spotify for music and YouTube for videos. Additionally, location-based
services using Open StreetMap are employed to display nearby psychologists, ensuring
comprehensive support for users
Conclusion:
In conclusion, the development of a web application for depression assessment and support
presents a significant advancement in addressing mental health challenges in a user-centric
manner. By integrating various modules such as initial assessment, depression severity analysis,
Future Scope:
Looking ahead, there are several promising avenues for expanding the capabilities and reach of
the depression assessment and support platform. One potential area for future development lies
in the integration of wearable devices to gather real-time physiological data, such as heart rate
variability and sleep patterns, which can provide valuable insights into an individual's mental
health status. By incorporating data from wearables, the platform can offer more personalized
and dynamic assessments, enabling timely interventions and support.
Furthermore, the creation of an Android application would extend the accessibility of the
platform to a broader audience, allowing users to conveniently access its features on their
mobile devices. With the widespread use of smartphones, an Android application would
enhance the reach and usability of the platform, particularly among users who prefer mobile-
based solutions.
Additionally, there is potential for the platform to serve as a centralized hub connecting users
with mental health professionals, including psychologists and counsellors. By leveraging
technology to facilitate online meetings and consultations, the platform can provide users with
easy access to professional support services, regardless of their geographical location. A
centralized platform connecting psychologists can also streamline communication and
collaboration among mental health professionals, fostering a supportive community and
ensuring comprehensive care for users.
Incorporating these future enhancements would not only enhance the functionality and usability
of the platform but also contribute to its effectiveness in supporting individuals' mental well-
being. As technology continues to evolve and mental health awareness grows, there is
significant potential for the platform to play a pivotal role in promoting mental wellness and
providing accessible support to those in need.
2. Title: IOT and ML Driven Smart Fire Alarm and Crowd Tracking.
Introduction
In recent years, fire-related injuries have caused widespread devastation and significant
loss of life. Fires in buildings are often unplanned or accidental, exacerbated by dense
residential development, tangled wires, and inadequate fire safety standards.
There are two main types of fire detection methods: traditional and vision-based systems.
Traditional methods use sensors like thermal, flame, and smoke sensors, but they struggle to
detect dynamic fire features such as motion and color. They also lack multidimensional data
on fire and smoke, like location and severity, requiring human intervention to trigger alarms
and confirm fire.
Vision-based systems are promising for fire detection, using cameras to detect static and
dynamic fire features like color and motion. They process information through light emission
and reflection, offering real-time visual data recording. These systems can also identify the
number of people trapped in a building during a fire, crucial when visibility is reduced due to
smoke and flame.
Current traditional fire alarm systems have limitations in surveillance coverage, reaction
time, human interaction, and detailed fire information. Complexity and false triggering
remain concerns. Dynamic exit path guidance during a fire is critical, where a dynamic exit
path can be more effective than a static one.
In a recent study, a Raspberry Pi-based vision system was implemented for detecting smoke
flow and estimating the number of people using a deep neural network (DNN). The DNN
model was trained with a sample dataset and used in real-time for fire detection and people
counting. Additionally, a web application was designed to visualize real-time visual data from
the vision system.
1. System Architecture :
As shown in the figure below, the input can be a text paragraph or document. The given input
will through the following stages : pre-processing i.e data cleaning and normalization,
processing i.e unique matrix generation, word count per sentences, frequency list of
sentences and calculating TFIDF matrix and post-processing i.e cosine similarity matrix
generation, graph generation and score generation using Textrank algorithm and sentence
ranking. During this process, some NLP methods will also be used to enhance the output.
After all the processing, the summarized output will be generated.
2. Conclusion:
The project offers a comprehensive solution for fire detection and people counting,
addressing safety and security concerns in crowded public spaces. Its real-world
applications extend to disaster management, industrial safety, and IoT integration, with the
potential for adaptive learning and ethical considerations. The project's adaptability and
customization make it a valuable tool in enhancing public safety and security across
diverse environments.
Future Scope:
Cloud-Based Solutions: Integration with cloud computing platforms for scalable data
storage, analysis, and remote access, facilitating more comprehensive monitoring of
multiple sites. Augmented Reality (AR) Integration: Augmented reality applications
for real-time visualization of crowd movement and emergency response coordination.
Public Safety and Security: The project's real-time fire detection and people counting
capabilities can be integrated into smart cities and public spaces to enhance safety and
security, making it valuable for large gatherings, transportation hubs, and more.
Customization for Different Environments: The system's adaptability can be harnessed
for various environments, from urban settings to remote areas, with customization to
suit specific needs.
Group members:
Bhushan Kadam
Gautam Patil
Vaishnavi Patil
Vedika Gujalwar
Guide Name: Dr.K.S.Wagh
Project Sponsored By: Bright Infotech
Introduction
KeyFeatures:
Advanced face detection: The system uses the MTCNN algorithm for accurate and
efficient face detection in input video streams. MTCNN uses a multi-level convolutional
neural network to detect facial regions, which ensures robust detection even in
challenging lighting conditions and various positions.
Facial Feature Extraction: After face detection, our system uses ResNet50, a state-of-
the-art deep neural network architecture, to extract high-value features from facial
images. ResNet50 is known for its ability to capture discriminative features with
remarkable accuracy, allowing accurate identification of individuals.
Classification: After feature extraction, the system uses a classification layer to classify
detected faces into predefined categories. Using the power of deep learning, our system
can accurately identify known individuals and distinguish them from unknown subjects.
Scalability and Efficiency: Our proposed system is designed to be highly scalable and
efficient, allowing seamless integration into existing surveillance infrastructure. It can
process large amounts of video data in real time, making it suitable for deployment in
various environments such as airports, commercial enterprises and public spaces.
Adaptive Learning: To improve performance over time, our system includes adaptive
learning mechanisms that constantly update and refine the underlying models based on
new data. This adaptive approach ensures that the system remains robust and effective
in dynamic environments with evolving facial characteristics.
Conclusion
In summary, the creation and deployment of a sophisticated human monitoring system that
incorporates facial recognition algorithms marks a noteworthy progression in the domains
of security, access management, and other related fields spanning numerous sectors. This
technology allows for the tracking and analysis of people’s movements and actions, as well
as real-time monitoring and automated alarm production. Even though improved security,
productivity, and data insights have many benefits, it is crucial to handle the drawbacks—
such as privacy issues, inconsistent accuracy, and legal issues with responsible use and
ethical considerations.
Future scope
1. Advancements in machine learning and deep learning techniques will likely lead to
improved accuracy.
3. Combining face recognition with other biometric modalities, such as iris or fingerprint
recognition.
Group members:
Ishan Mankar
Amogh Gojamgunde
Niraj Bondarde
Harshada Gaikwad
Introduction:
In light of the prevailing challenges associated with parking management in several
metropolitan areas, there arises a pressing need for a sophisticated automated system aimed at
not only efficiently managing parking but also reducing the reliance on manual labor.
The application will harness Advanced Number Plate Recognition (ANPR) technology to
accurately identify vehicles, encompassing non-standard and decorative number plates. It will
systematically register vehicles in a centralized database, complete with timestamp records,
offer recommendations for available parking spaces, and facilitate the allocation of parking
spots upon vehicle entry.
In order to guarantee practical utility, a dedicated web application has been meticulously
crafted. This platform empowers parking owners to visualize real-time data collected by the
parking management system, enabling them to gain insights into their sales metrics, while
simultaneously providing users with timely notifications regarding parking space availability.
Proposed System:
Correctly positioned parking prompts precise fee calculation, considering entry and exit
times. Simplifying transactions, the system generates QR codes, culminating in the issuance
of comprehensive digital receipts.The proposed system model is depicted in figure 1.
Key features include:
1. User Authentication and Registration: The system initiates with user authentication and
registration processes, ensuring secure access to the platform.
2. PayWallet Integration: Upon successful login, users can conveniently add funds to their
PayWallet for streamlined payments, enhancing transaction efficiency.
3. Parking Space Reservation: Users can reserve parking spaces through the system, which
validates bookings and deducts charges directly from their PayWallet, ensuring smooth
transaction processing.
4. License Plate Scanning: Smooth entry is
facilitated through license plate scanning, with the
system verifying plate registration status to ensure
compliance.
5. Vehicle Details Input: Unregistered plates
prompt users to input vehicle details, securely
stored post-verification, enhancing data integrity
and security.
6. Parking Space Suggestions: For registered
plates, the system suggests available parking
spaces, optimizing parking space utilization and
user convenience.
7. Incorrect Parking Monitoring: The system actively monitors for incorrect parking
positions, promptly notifying both users and administrators upon detection, ensuring adherence
to parking regulations.
8. Fee Calculation: Correctly positioned parking prompts precise fee calculation, considering
entry and exit times, facilitating transparent and fair pricing.
9. QR Code Generation: Simplifying transactions, the system generates QR codes for
seamless payment processing, enhancing user convenience and accessibility.
10. Digital Receipt Issuance: Upon completion of transactions, comprehensive digital receipts
are issued, providing users with detailed records of their parking activities.
Conclusion:
In this integrated approach, we are implementing a parking management system that makes use
of ANPR to recognize all plate types, register vehicles upon entry, provide real-time parking
recommendations, enable advance parking reservations, assign parking spaces post-entry,
facilitate various payment methods, and generate comprehensive receipts. This approach aims
to enhance parking management by harnessing ANPR technology, optimizing space allocation,
simplifying payment processes, and offering the convenience of advance reservations.
Future Scope:
Looking ahead, A user-friendly map feature will provide step-by-step directions, guiding users
back to their parked cars seamlessly. In addition, an information center section will be
seamlessly integrated to offer comprehensive insights into nearby service hubs, encompassing
petrol pumps, EV charging stations, garages, and car wash facilities, enhancing the overall user
experience. Future implementation will consist of algorithms to detect unusual group activities,
such as fights or violent behavior, The system will alert the relevant authorities in real-time for
prompt action. Additionally, it will detect individual activities like damaging a parked car or
theft attempts and trigger appropriate alerts. Fire detection algorithms will also be integrated to
identify potential fire hazards and respond quickly using .
Our forward-looking will introduce additional features to enrich user experience by introducing
the option to reserve parking spaces in advance through the app, making it convenient for
special events or high-demand periods. Rental parking spaces will also be available for
overnight parking or monthly subscriptions, providing flexibility for long-term parking needs.
With its foundation in ANPR technology and multifaceted features, it sets forth new
benchmarks in the dynamic smart parking industry. We user-centric, and secure aspects that
make our parking management system a pioneering solution for the modern urban environment.
Group Members:
Salonee Pathan.
Hardik Kotangale
Rahul Rajpurohit
Atharva Tirkhunde
Introduction:
The convergence of blockchain technology and electronic voting systems has led to the
emergence of innovative and secure methods for conducting transparent and tamper-resistant
elections. The proposed system delves into the design and implementation of an E-voting
system utilizing the robustness of blockchain technology, with a specific focus on the utilization
of the SHA-256 algorithm for enhanced security. The proposed E-voting system utilizes
blockchain's characteristics to tackle key issues found in traditional voting systems like voter
anonymity, ballot integrity, and the ability to verify votes. Cryptographic techniques are
employed, and the SHA-256 algorithm is crucial in maintaining the immutability and integrity
of data within the blockchain. This, in turn, ensures the legitimacy of cast votes and the overall
integrity of the election process. It establishes a self-governing environment through the
application of blockchain's consensus mechanisms, specifically Proof of Work (PoW). This
autonomy allows participants to confidently engage in the electoral process without reliance on
central authorities. The inherent transparency of the blockchain ensures that each vote is
publicly accessible while maintaining the privacy of individual voters through cryptographic
techniques. Additionally, the inclusion of smart contracts enhances automation and precision in
the electronic voting process by executing predefined rules and protocols and eliminating the
need for intermediaries, reducing the risk of human error and manipulation, streamlining the
voting procedure, ensuring consistency, and improving the accessibility of the electoral system.
System Architecture:
The E-voting system integrates blockchain technology to ensure a secure and transparent voting
process. Initially, the system requires voters and candidates to register, during which personal
information such as name, identification, and contact details are collected. This phase also
involves the generation of RSA key pairs, with the public key being stored in the system’s
database and the private key retained by the participant. Once registration is complete, identity
verification is carried out through a challenge-response mechanism using the stored public keys.
Participants prove their identity by decrypting a challenge encrypted with their public key using
their private key. Successful decryption and response submission confirm the participant’s
identity. The core of the system lies in its smart contracts, which encapsulate the voting logic.
These contracts verify voter eligibility based on criteria like citizenship, age, and registration
status. They enable securevoting, prevent double voting, and ensure the confidentiality and
anonymity of each vote. Election setup is another critical step, where smart contracts define the
election’s parameters, such as type, date, and eligible participants, with all details recorded on
the blockchain. During the voting process, voters encrypt their votes using RSA encryption and
attach a digital signature created with their private key. This combination of encryption and
digital signatures maintains the confidentiality and integrity of the vote. Finally, the vote
tallying module comes into play. The system verifies the digital signatures to authenticate each
vote. Authorized parties then decrypt the votes, and smart contracts automatically tally them.
The results are published and recorded on the blockchain, concluding the process with a secure,
accurate, and tamper-proof election. The e-voting system project holds vast potential for the
future. It can evolve by incorporating advanced Blockchain technologies, biometric
authentication, making it more secure and accessible. Research into global adoption, quantum
computing resilience, and improved accessibility features will broaden its impact. Public
awareness campaigns, legal compliance, and public-private partnerships are vital for its
success.In the coming years, the e-voting system project can evolve to embrace cutting-edge
innovations. Enhanced user experiences, seamless integrations with emerging technologies, and
adaptive security measures will be key focuses. The continuous pursuit of transparency,
accessibility, and adaptability will position the project to be a cornerstone in the ongoing
transformation of digital democracy and secure voting practices.
Conclusion:
simplicity and ease of use, encouraging widespread adoption and participation. By addressing
the limitations of conventional voting systems, it offers a solution that is not only secure and
transparent but also readily accessible to voters. The implementation of a custom blockchain,
provides a robust platform that can handle large volumes of data while maintaining quick
transaction times. This innovative approach to voting paves the way for a future where
democratic processes are more inclusive, efficient, and resilient against fraud, making it
apivotal development in the evolution of electoral systems.
Future Works:
The e-voting system project holds vast potential for the future. It can evolve by incorporating
advanced Blockchain technologies, biometric authentication, making it more secure and
accessible. Research into global adoption, quantum computing resilience, and improved
accessibility features will broaden its impact. Public awareness campaigns, legal compliance,
and public-private partnerships are vital for its success. In the coming years, the e-voting
system project can evolve to embrace cutting-edge innovations. Enhanced user experiences,
seamless integrations with emerging technologies, and adaptive security measures will be key
focuses. The continuous pursuit of transparency, accessibility, and adaptability will position the
project to be a cornerstone in the ongoing transformation of digital democracy and secure
voting practices.
Group Members
Falguni Takawale
Aarti Madane
Sanskruti Pawar
Tazeen Sayed
Guide Name:Mrs. Archana Said.
Title: Neuroscan Parkinson Detection Using Voice, Handwriting and Gait Dataset
Introduction:
With many benefits and the potential to significantly improve people's health and wellness, the
Parkinson's Disease Detection Web Application is a promising and cutting-edge method for the
early diagnosis of Parkinson's disease. With the help of this ground breaking app, users can
easily and non-invasively determine their risk of developing Parkinson's disease, increasing
their sense of awareness and control. Through the combination of spiral drawing , voice sample
analysis and gait dataset ,the application enhances diagnostic accuracy, providing a more
comprehensive assessment of disease risk. Users are presented with clear and understandable
prediction results, enabling them to make informed healthcare decisions. The user-friendly
interface ensures accessibility for individuals with varying technical abilities, promoting
inclusivity. However, it is essential to recognize the limitations and dependencies of the
application. Data quality, ethical data usage, and privacy concerns should be carefully
addressed. The application should not replace clinical diagnosis but rather serve as a
supplementary tool for early detection and self-assessment .The potential applications of this
tool are vast, spanning from individual self-assessment to healthcare support, research,
telemedicine, public health initiatives, and more. It can contribute to raising awareness,
advancing research, and promoting ethical data usage in healthcare. To sum up, the Parkinson's
Disease Detection Web Application exemplifies the spirit of technological innovation in
healthcare with the goal of improving early detection and enabling people to take proactive
measures towards improved health. Notwithstanding its drawbacks and difficulties, it is a useful
addition to the field of Parkinson's disease assessment and study because of its potential
advantages for early detection, user involvement, and public health
System Architecture:
The proposed Parkinson's disease detection system comprises a user-friendly web interface
allowing seamless data upload and result presentation, ensuring accessibility across user
demographics. By amalgamating handwriting, voice, and gait data analyses, the system
employs convolutional neural networks (CNN) for nuanced handwriting assessment, machine
learning models for precise voice analysis, and support vector machine (SVM) classifiers for
comprehensive gait examination. Emphasizing heuristic principles, the interface prioritizes
simplicity and effectiveness, aiming to provide accurate diagnosis and monitoring capabilities.
Ultimately, the system endeavors to enhance patient care and satisfaction by offering reliable
and accessible Parkinson's disease detection.
Conclusion:
With many benefits and the potential to significantly improve people's health and well-being,
the Parkinson's Disease Detection Web Application is a promising and cutting-edge method for
the early diagnosis of Parkinson's disease. With the use of this innovative software, users may
easily and non invasively determine their risk of developing Parkinson's disease, increasing
their sense of awareness and control .Through the combination of spiral drawing and voice
sample analysis, the application enhances diagnostic accuracy, providing a more comprehensive
assessment of disease risk. Users are presented with clear and understandable prediction results,
enabling them to make informed healthcare decisions. The user friendly interface ensures
accessibility for individuals with varying technical abilities, promoting inclusivity. However, it
is essential to recognize the limitations and dependencies of the application. Data quality,
ethical data usage, and privacy concerns should be carefully addressed. The application should
not replace clinical diagnosis but rather serve as a supplementary tool for early detection and
self assessment. The potential applications of this tool are vast, spanning from individual self-
assessment to healthcare support, research, telemedicine, public health initiatives, and more. It
can contribute to raising awareness, advancing research, and promoting ethical data usage in
healthcare. To sum up, the Parkinson's Disease Detection Web Application exemplifies the
spirit of technological innovation in healthcare with the goal of improving early detection and
enabling people to take proactive measures towards improved health. It is a useful addition to
the field of Parkinson's disease assessment and study, despite its limitations and difficulties.
These advantages may include early diagnosis, user engagement, and public health.
Group Members:
Yash Paraswar
.Sanket Mane
Sanket Raut
Madhura Urunkar