0% found this document useful (0 votes)
19 views6 pages

Project On Big Data

The document describes a project to generate metadata for ISRO mission videos using deep learning. It aims to automate the labor-intensive metadata process and improve video classification accuracy. The system will categorize videos and generate metadata like titles, descriptions, timestamps to facilitate search and retrieval of ISRO's video collection.

Uploaded by

Hemanth krishna
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
19 views6 pages

Project On Big Data

The document describes a project to generate metadata for ISRO mission videos using deep learning. It aims to automate the labor-intensive metadata process and improve video classification accuracy. The system will categorize videos and generate metadata like titles, descriptions, timestamps to facilitate search and retrieval of ISRO's video collection.

Uploaded by

Hemanth krishna
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Project Team #: 20CSM_B06

20BQ1A4259 – V. Mohan
20BQ1A4251 – S. Sai Kumar
20BQ1A4256 – T. Hemanth Krishna
20BQ1A4226 – K. Sai Venkata Phani Bhargav

Project Name:
Video Metadata Generation and Classification using Deep Learning Techniques.

Abstract:
The project aims to enhance the efficiency of video metadata generation for ISRO mission and
program documentaries by automating the labor-intensive process. The existing approach employs a
hybrid model of CNN and LSTM, but it faces challenges related to overfitting and high computational
cost. To address these issues, the proposed solution focuses on improving the accuracy of video
classification through meticulous fine-tuning and hyperparameter optimization. The system's primary
objective is to effectively categorize videos into diverse genres, such as launch programs, interviews,
educational content, outdoor shots, public events, traffic, and more. By generating metadata like Title,
Description, Timestamps, Duration, Tags, and Keywords, the system aims to analyze video content
efficiently and extract pertinent information. This metadata will facilitate easy classification and
retrieval, ultimately enhancing the accessibility and utility of ISRO's extensive collection of video
documentaries.
Title Video Metadata Generation and Classification using Deep
Learning Techniques

Client Department of Space, Indian Space Research Organisation


(ISRO).

Objective The main objective of the proposed solution is to streamline


and automate the labor-intensive process of video metadata
generation for ISRO mission and program documentaries.
The system aims to improve the accuracy of video
classification by carefully fine-tuning and optimizing
hyperparameters. By generating metadata like Title,
Description, Timestamps, Duration, Tags, and Keywords,
the solution seeks to efficiently analyze video content and
extract pertinent information. This will enable the system to
effectively categorize videos into diverse genres, including
launch programs, interviews, educational content, outdoor
shots, public events, traffic, and more. The ultimate goal is
to enhance the accessibility and utility of ISRO's extensive
collection of video documentaries, making it easier for users
to classify, search, and retrieve relevant videos based on their
content and characteristics.
Users 1. Employees of the Department of Space.
2. Video Archivists and Curators.
3. Researchers and Analysts.
4. Educators and Trainers.

Functional 1. User Authentication and Account Management:


Requirements a. This requirement ensures secure user registration,
login, and password reset functionalities.
b. It includes user authentication mechanisms like
email verification and password encryption to
safeguard user data.
c. User authentication is critical for data privacy and
access control, ensuring that only authorized
personnel can use the application and manage videos.
2. Video Upload and Management:
a. This functionality allows users to upload videos
related to ISRO missions and programs.
b. It involves efficient storage and retrieval
mechanisms to manage the increasing volume of
uploaded videos.
c. Video management ensures that users can easily
organize and access their uploaded videos, including
features like viewing, editing, and deleting videos.
3. Video Classification:
a. The video classification feature allows users to
categorize uploaded videos into relevant genres or
categories, such as mission type, program theme, or
event type.
b. Proper video classification enhances video
organization and retrieval, enabling users to find
specific videos quickly based on their categories.
4. Video Search:
a. The search functionality enables users to find videos
by entering relevant keywords or video titles in the
search bar.
b. Video search streamlines the process of locating
specific videos among the vast collection, saving
time and improving user experience.

5. Video Statistics and Dashboard:


a. Providing a dashboard with video statistics allows
users to view insights into the performance and usage
of their uploaded and classified videos.
b. Video statistics, such as view counts, genre
distribution, and user engagement metrics, help users
gauge the popularity and impact of their videos.

6. User Roles and Permissions:


a. Implementing different user roles (e.g., admin,
moderator, regular user) ensures varying levels of
access and control over application features and data.
b. User roles and permissions are crucial for data
security and preventing unauthorized access to
sensitive functionalities.

7. Video Sharing and Collaboration:


a. Enabling users to share videos with other authorized
users fosters collaboration and knowledge sharing
within the organization.
b. Video sharing promotes teamwork and enables
efficient dissemination of information among
different teams and stakeholders.

8. Video Playback and Streaming:


a. Providing smooth video playback and streaming
capabilities ensures a seamless viewing experience
for users.
b. Reliable video playback and streaming support
enhance user satisfaction and accessibility of video
content across various devices and network
conditions.
Non-functional 1. Security and Privacy:
Requirements a. This requirement focuses on ensuring the
application's security to protect sensitive user data
and prevent unauthorized access to videos and user
accounts.
b. Robust encryption mechanisms, secure
authentication, and role-based access control should
be implemented to safeguard user information and
video content.

2. Scalability:
a. The application should be able to handle a large
number of users and video uploads without
compromising performance.
b. Designing the application with scalability in mind
ensures that it can accommodate growing user bases
and increased video content.

3. Reliability and Availability:


a. The application must be reliable, minimizing
downtime and ensuring that users can access their
videos and perform tasks consistently.
b. Implementing redundant systems, load balancing,
and automated backups can enhance the application's
availability and reliability.

4. Performance:
a. The application should deliver fast response times for
video uploads, classification, and searches, providing
users with a smooth and responsive experience.
b. Optimizing database queries, caching frequently
accessed data, and leveraging efficient algorithms
contribute to improved application performance.

5. User Experience (UX):


a. The application's user interface should be intuitive,
user-friendly, and visually appealing to enhance the
overall user experience.
b. Conducting usability testing and gathering user
feedback during development can lead to an
application that meets user expectations and
preferences.

6. Compatibility and Cross-Platform Support:


a. The application should function seamlessly across
different devices, operating systems, and web
browsers.
b. Ensuring compatibility with various platforms
extends the application's reach and accessibility to a
wider audience.

7. Data Backup and Recovery:


a. Regular and automated data backups are essential to
prevent data loss in case of hardware failures or other
unforeseen incidents.
b. Implementing reliable data backup and recovery
mechanisms provides data integrity and minimizes
the impact of potential disasters.

8. Regulatory Compliance:
a. The application should adhere to relevant regulations
and data protection laws to ensure legal compliance.
b. Complying with data privacy regulations and any
specific industry standards strengthens user trust and
protects the organization from legal issues.

Software and Hardware Software Requirements:


Requirements 1. Integrated Development Environment (IDE):
 Visual Studio Code.
2. Front-end Technologies:
 HTML, CSS, and JavaScript.
 React.js.
3. Back-end Technologies:
 Node.js.
 Express.js.
 Django.
4. Databases:
 MySQL.
 MongoDB.
5. User Authentication:
 Passport.js.
6. Video Processing and Metadata Extraction:
 FFmpeg
7. Video Classification:
 TensorFlow.js
8. Deep Learning Framework:
 TensorFlow or PyTorch for implementing CNN and
LSTM models
9. Cloud Storage:
 Amazon S3
10. Content Delivery Network (CDN):
 Amazon CloudFront
Hardware Requirements:
1.Development Machine:
 Processor: A modern multicore processor (e.g., Intel
Core i7 or AMD Ryzen 7) with a clock speed of at
least 3.0 GHz or higher.
 RAM: 16GB to 32GB of RAM.
 Storage: An SSD (Solid State Drive) with at least
500GB of storage space.
 Graphic Card: A dedicated graphics card (e.g.,
NVIDIA GeForce GTX or RTX series) with at least
4GB of VRAM.
 Operating System: Windows 10, macOS, or a
popular Linux distribution (e.g., Ubuntu, CentOS).

PROJECT GUIDE HOD- CSM

You might also like