Stack Tracer: Dual-Phase Steganalysis and Malware Detection for

Multimedia Security

K. Rithik Karthikeyan

, M. B. Manav Srinivas

and A. Jenifer

Artiﬁcial Intelligence and Data Science, St. Joseph’s Institute of Technology, Chennai, India

Keywords:

Steganalysis, Hidden Data Detection, Multimedia Security, Browser Extension, VirusTotal Integration,

Malware Analysis, Technical Analysis Tools, Real-Time Threat Assessment, Cybersecurity, Data Hiding

Techniques.

Abstract:

The increasing use of steganographic techniques to embed hidden and potentially harmful data within mul-

timedia ﬁles poses signiﬁcant challenges to cybersecurity. Existing detection methods often lack precision,

scalability, and real-time capabilities, necessitating innovative solutions. This paper introduces Stack Tracer, a

browser extension designed to detect and analyze hidden data across various media formats, including images,

audio, video, and more. The extension integrates multiple advanced technical analysis tools into a uniﬁed

backend platform for detecting concealed content. Detected results are seamlessly passed to VirusTotal’s

API, enabling comprehensive threat assessment by leveraging its extensive malware database. Unlike tra-

ditional methods, Stack Tracer validates its contributions through detailed comparisons with state-of-the-art

tools, demonstrating detection rates of 95% for images, 92% for audio, and 89% for video. These results

establish its robustness and accuracy across diverse media types. With a clear problem deﬁnition, validated

outcomes, and a user-friendly interface, Stack Tracer provides a reliable and accessible tool for real-time mul-

timedia threat analysis, addressing gaps in existing solutions.

1 INTRODUCTION

The growing complexity of multimedia content in the

digital landscape has led to an increase in the use

of steganographic techniques to embed hidden data

within various media ﬁles such as images, audio, and

video. While steganography has legitimate applica-

tions, it is increasingly exploited for malicious activ-

ities, including unauthorized data breaches and mal-

ware distribution (Shehab and Alhaddad, 2019; Kaur

and Behal, 2020). Traditional detection methods of-

ten fall short due to their fragmented nature, focusing

on speciﬁc media types or relying on manual analy-

sis. These approaches lack scalability and real-time

applicability, making it difﬁcult to provide compre-

hensive and timely protection (Abdelfattah and Mah-

mood, 2021; Subramanian et al., 2020). Furthermore,

many existing tools fail to integrate threat validation,

leaving users without clear insights into whether the

detected hidden data could pose any security risks

https://orcid.org/0009-0000-4391-1451

https://orcid.org/0009-0006-9817-988X

https://orcid.org/0009-0003-7967-2175

(Fridrich, 2019).

To overcome these limitations, Stack Tracer is in-

troduced—a browser extension that integrates mul-

tiple steganalysis tools to detect hidden data across

various media formats. By automating the detection

process, Stack Tracer simpliﬁes the identiﬁcation of

embedded content and passes the results to VirusTo-

tal’s API for real-time threat assessment. VirusTotal’s

extensive malware database is leveraged to evaluate

whether the detected data is harmful, providing users

with actionable insights (Takao et al., 2017; Virus-

Total, 2024b). This dual-phase detection framework

not only enhances accuracy but also offers a seamless,

user-friendly interface that makes sophisticated back-

end processes accessible to end-users.

Through extensive testing, Stack Tracer has

achieved detection rates of 95% for images, 92% for

audio, and 89% for video, validating its effective-

ness in handling multiple media types while maintain-

ing high accuracy and reliability. Additionally, Stack

Tracer ﬁlls a critical gap in existing solutions by offer-

ing real-time, user-friendly protection against hidden

threats in digital content.

890

Karthikeyan, K. R., Srinivas, M. B. M. and Jenifer, A.

Stack Tracer: Dual-Phase Steganalysis and Malware Detection for Multimedia Security.

DOI: 10.5220/0013606700004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 890-896

ISBN: 978-989-758-763-4

2 RELATED WORKS

The detection of hidden data within multimedia ﬁles

and its subsequent threat analysis is a critical chal-

lenge in modern cybersecurity. Numerous research

efforts have explored various aspects of steganogra-

phy, steganalysis, and malware detection. This sec-

tion reviews key contributions that form the foun-

dation for the design and development of the Stack

Tracer browser extension.

Shehab and Alhaddad (Shehab and Alhaddad,

2019) conducted an extensive survey on multimedia

steganalysis, highlighting the effectiveness of statis-

tical methods such as histogram analysis and spatial

rich modeling (SRM) in detecting hidden data. Their

ﬁndings provide valuable insights into the design of

Stack Tracer’s image analysis framework. Similarly,

Hameed et al. (Hameed et al., 2020) reviewed ad-

vanced steganographic techniques and their counter-

measures, emphasizing the need for robust algorithms

to detect concealed content across diverse media for-

mats.

Fridrich (Fridrich, 2019) outlined foundational

methodologies in steganalysis, focusing on the in-

tegration of statistical and machine-learning ap-

proaches. This work serves as a basis for combining

multiple detection tools in Stack Tracer. Subrama-

nian et al. (Subramanian et al., 2020) explored recent

advancements in image steganography and detection

techniques, reinforcing the importance of integrating

state-of-the-art methods into comprehensive systems

like the proposed extension.

Kaur and Behal (Kaur and Behal, 2020) provided

a detailed review of text-based steganography tech-

niques, which is relevant to extending the capabilities

of Stack Tracer beyond traditional image and video

formats. Meghanathan and Nayak (Meghanathan and

Nayak, 2021) examined multimodal detection tech-

niques, showcasing the necessity of systems capable

of handling multiple media types, a key feature of

Stack Tracer.

Meg

ıas et al. (Meg

ıas et al., 2020) investigated

data-hiding methods and watermarking techniques in

multimedia, highlighting the challenges of detecting

subtle patterns in hidden data. Their ﬁndings sup-

port the need for advanced analysis tools integrated

into platforms like Stack Tracer. Abdelfattah and

Mahmood (Abdelfattah and Mahmood, 2021) dis-

cussed emerging trends in steganography, emphasiz-

ing the need for scalable and adaptable solutions,

which aligns with the extension’s goal of providing

real-time, multi-format support.

Takao et al. (Takao et al., 2017) proposed a frame-

work leveraging VirusTotal for URL-based threat

analysis, demonstrating the utility of integrating ex-

ternal databases for comprehensive security evalua-

tions. This concept is a cornerstone of Stack Tracer,

which uses VirusTotal to validate the safety of de-

tected hidden content. Turner and Lee (Turner and

Lee, 2022) reviewed real-time malware detection

techniques, underscoring the importance of seamless

threat assessment, a feature central to the extension.

The reviewed works collectively address various

facets of steganography and threat analysis but often

lack a uniﬁed framework for detecting hidden data

and evaluating its potential risks. Stack Tracer bridges

these gaps by integrating multiple advanced analysis

tools into a user-friendly browser extension, offering

comprehensive detection and real-time threat valida-

tion across diverse media types.

3 TOOLS AND METHODOLOGY

The Stack Tracer framework uses traditional and

deep learning-based steganalysis techniques to detect

hidden data in multimedia ﬁles (Fridrich, 2019).

3.1 Steganalysis Techniques

To effectively detect hidden data in images, videos,

audio ﬁles, and URLs, Stack Tracer integrates tra-

ditional steganalysis tools with deep learning tech-

niques.

3.1.1 Traditional Steganalysis Methods

Various statistical and heuristic-based tools are used

to identify anomalies that indicate steganographic em-

bedding.

Image Steganalysis Tools such as StegExpose,

Stegdetect, and StegoSuite analyze pixel value

changes, statistical distributions, and noise patterns

(Fridrich, 2019). The detection accuracy is based on

statistical deviation σ from expected pixel values:

image



original

− P

stego

original



× 100 (1)

where P

original

and P

stego

represent pixel distribu-

tions in original and modiﬁed images.

Video Steganalysis The VideoStegAnalyzer tool is

used to examine frame differentials and motion vec-

tor inconsistencies to detect steganographic embed-

ding (Meghanathan and Nayak, 2021).

Stack Tracer: Dual-Phase Steganalysis and Malware Detection for Multimedia Security

891

Audio Steganalysis StegAlyzerAS is used to de-

tect hidden data in audio ﬁles by analyzing frequency

shifts and phase distortions. The signal-to-noise ratio

(SNR) is calculated as:

SNR = 10 log



signal

noise



(2)

where P

signal

and P

noise

denote the power levels of

the actual signal and suspected hidden data (Kaur and

Behal, 2020).

URL Steganalysis URL steganalysis detects hid-

den data within shortened links, query parameters, or

domain names by analyzing entropy levels and meta-

data (Browsing, 2022; VirusTotal, 2024b). The en-

tropy H of a URL string is calculated as:

H = −

∑

i=1

p(x

)log

p(x

) (3)

where p(x

) represents the probability of each

character in the URL.

3.1.2 Deep Learning-Based Steganalysis

Traditional tools may struggle against adaptive

steganography techniques. To enhance detection ac-

curacy, deep learning models are integrated.

CNN-based Image Steganalysis Xu-Net, a convo-

lutional neural network (CNN) model, extracts fea-

tures from images to differentiate between normal and

steganographic content (Hameed et al., 2020). The

classiﬁcation conﬁdence C is calculated as:

C =

∑

i=1

softmax( f (x

, W )) (4)

where N is the total number of image samples,

represents feature vectors, and W denotes model

weights.

Vision Transformer (ViT) for Image Steganalysis

Vision Transformers (ViTs) process entire images to

detect hidden data by analyzing spatial inconsisten-

cies (Smith and Johnson, 2022). The anomaly score

is deﬁned as:

= ∥F

ViT

(I) − I

clean

∥ (5)

where F

ViT

(I) represents the processed image fea-

tures.

3.2 Threat Analysis and Malware

Detection

Extracted content is assessed for potential malware

threats.

3.2.1 VirusTotal API Integration

Extracted content is sent to VirusTotal, which scans

the ﬁle against multiple antivirus engines (Takao

et al., 2017; VirusTotal, 2024a). The threat score is

computed as:

T =

× 100 (6)

where M is the number of antivirus engines ﬂag-

ging the content, and E is the total number of engines.

3.2.2 Sandbox Analysis

If VirusTotal results are inconclusive, the extracted

content is executed in a controlled sandbox environ-

ment to observe its behavior (Kaspersky, 2022a).

4 PROPOSED SYSTEM

The proposed system, Stack Tracer, addresses the

limitations of existing technologies by providing a

uniﬁed, real-time, and scalable solution for multime-

dia security analysis. This section details the archi-

tecture, workﬂow, and features of the system, high-

lighting its contributions to overcoming current chal-

lenges.

4.1 Existing Technologies and

Limitations

Existing steganalysis tools and platforms such as

VirusTotal have made signiﬁcant progress in detect-

ing hidden data and assessing threats. However, they

face the following limitations:

• Single Media Type Focus: Many steganalysis

tools specialize in one type of media, such as im-

ages or audio, but lack adaptability to handle di-

verse formats like videos and URLs (Subramanian

et al., 2020), (Fridrich, 2019), (Shehab and Alhad-

dad, 2019).

• Lack of Integration: Tools like VirusTotal are

highly effective for malware analysis but require

manual input and cannot independently analyze

hidden data within multimedia ﬁles (Takao et al.,

2017), (Abdelfattah and Mahmood, 2021).

INCOFT 2025 - International Conference on Futuristic Technology

892

• Fragmented Workﬂow: Current systems lack a

uniﬁed interface for steganalysis and threat as-

sessment, leading to inefﬁciencies and increased

complexity for users (Hameed et al., 2020), (Kaur

and Behal, 2020).

• Limited Real-Time Functionality: Few exist-

ing solutions are optimized for real-time anal-

ysis and scalability, making them unsuitable

for high-volume or time-sensitive applications

(Meghanathan and Nayak, 2021), (White and

Green, 2022).

4.2 System Architecture

The architecture of the Stack Tracer system integrates

multiple components to overcome the above limita-

tions. The design, shown in Fig. 1ﬁgure.1, consists

of:

• Browser Extension Interface: Captures mul-

timedia content directly from user interactions

within the browser.

• Content Script and Background Script: Facil-

itates communication between the user interface

and backend processing modules.

• Steganalysis Engine: Detects hidden data within

multimedia ﬁles, employing advanced method-

ologies such as multi-modal feature extraction

and deep-learning-based analysis (Meg

ıas et al.,

2020), (Arora, 2021).

• VirusTotal Integration: Automates API calls to

assess threats and provide comprehensive mal-

ware analysis results (Turner and Lee, 2022),

(Smith and Johnson, 2022).

• User Interface: Displays results in an interactive

dashboard with features such as history tracking,

phishing detection, and notiﬁcations (PhishTank,

2022), (Kaspersky, 2022a).

4.3 Workﬂow of the System

The workﬂow of the Stack Tracer system is illustrated

in Fig. 2ﬁgure.2. It consists of the following stages:

1. Content Capture: The browser extension cap-

tures multimedia content such as images, audio,

video, and text from web pages.

2. Steganalysis Detection: The captured content is

analyzed for hidden data using advanced steganal-

ysis techniques (Chrome, 2020), (Opera, 2020).

3. Threat Assessment: Detected data is sent to

VirusTotal for malware analysis and risk evalua-

tion (VirusTotal, 2024b).

Figure 1: Stack Tracer System Architecture

4. Result Delivery: The analysis results are visu-

alized on a user-friendly dashboard, highlighting

threats and providing actionable insights.

Figure 2: Workﬂow of the Stack Tracer Browser Extension

4.4 Features of the Proposed System

The Stack Tracer system incorporates the following

features:

• Real-Time Steganalysis: Detects hidden data in

Stack Tracer: Dual-Phase Steganalysis and Malware Detection for Multimedia Security

893

multimedia ﬁles across various formats (Brows-

ing, 2022).

• Integrated Threat Intelligence: Combines ste-

ganalysis detection with VirusTotal and Phish-

Tank for enhanced threat detection (Kaspersky,

2022b).

• User-Centric Design: Offers an intuitive dash-

board with history tracking, phishing detection,

and notiﬁcations for an improved user experience.

• Scalability: Optimized for high performance, en-

suring effective analysis of large datasets without

delays.

5 EXPERIMENTAL RESULTS

This section evaluates the performance of the Stack

Tracer system using multiple metrics, including De-

tection Accuracy, False Positive Rate (FPR), and Pro-

cessing Time. The evaluation spans diverse datasets

and provides insights into the system’s efﬁciency and

reliability. Additionally, a user interface visualization

is included to illustrate practical usability.

5.1 Dataset and Setup

The experiments utilized a curated dataset of 3,000

media ﬁles, including:

• Images: JPEG ﬁles with varying resolutions and

compression levels.

• Audio: WAV ﬁles encoded with different sam-

pling rates.

• Videos: MP4 ﬁles with diverse frame rates and

resolutions.

These ﬁles included both benign media and

steganographically altered samples. Hidden payloads

ranged from low-intensity steganographic data (im-

perceptible changes) to high-intensity data (easily de-

tectable distortions). This variety ensured a compre-

hensive evaluation of Stack Tracer.

The selection of datasets follows best practices

in multimedia forensics as outlined by White et al.

(White and Green, 2022). The authors highlight

the necessity of diverse datasets in steganalysis re-

search, emphasizing that both conventional and novel

steganographic techniques should be included to val-

idate system robustness.

The experiments were conducted on a machine

with the following speciﬁcations:

• Processor: Intel Core i7-9700K

• RAM: 16GB DDR4

• Environment: The backend integrated advanced

statistical analysis, frequency domain techniques,

and machine learning-based models for media

evaluation.

5.2 Evaluation Metrics

The following metrics were used for assessing the

system’s performance:

• Detection Accuracy: Measures the percentage of

correctly classiﬁed ﬁles:

Accuracy =

TP + TN

Total Samples

• False Positive Rate (FPR): Quantiﬁes how often

benign ﬁles are misclassiﬁed as harmful:

FPR =

FP + TN

• Processing Time: The average time required to

analyze a single ﬁle, highlighting the system’s

computational efﬁciency.

These metrics align with previous methodologies

used in steganalysis evaluations, such as those dis-

cussed in (White and Green, 2022; Fridrich, 2019),

ensuring that results can be compared to state-of-the-

art techniques.

5.3 Results and Analysis

The experimental outcomes are summarized in Ta-

ble 1table.1.

Table 1: Performance Metrics of Stack Tracer.

Metric Images (JPEG) Audio (WAV) Video (MP4)

Accuracy (%) 95 93 91

False Positive Rate (%) 4 5 6

Processing Time (s) 1.2 1.3 1.5

The system achieved an overall accuracy of 95%

for images, 93% for audio ﬁles, and 91% for video

ﬁles. It maintained a low false positive rate of 4% for

images, 5% for audio, and 6% for video. The aver-

age processing time for each ﬁle type highlights the

system’s computational efﬁciency, with a notable im-

provement over traditional methods.

The results indicate that Stack Tracer outperforms

existing tools in multimedia steganalysis. Similar

trends were observed in the study by White et al.

(White and Green, 2022), where models leverag-

ing hybrid approaches (statistical and machine learn-

ing) demonstrated superior detection rates in complex

datasets.

INCOFT 2025 - International Conference on Futuristic Technology

894

5.4 Visualization

To illustrate the real-world functionality of Stack

Tracer, Figure 3ﬁgure.3 presents a screenshot of the

system’s user interface while analyzing a JPEG ﬁle.

The interface provides users with options to upload

ﬁles, enter URLs, initiate analysis, and view histori-

cal results. In this example, the system has ﬂagged a

ﬁle (‘Untitled.jpeg‘) as harmful, displaying the detec-

tion result along with a timestamp for traceability.

Figure 3: User interface of Stack Tracer detecting a harmful

JPEG ﬁle. The system provides ﬁle status, timestamp, and

an option to review past analyses.

5.5 Comparative Analysis

To further validate the effectiveness of the Stack

Tracer system, its performance was compared with

existing tools used for steganalysis. The comparison

was based on key metrics such as detection accuracy,

false positive rate, and processing time. This allows

a clear understanding of Stack Tracer’s capabilities in

handling different multimedia formats. Table 2table.2

summarizes the results of this analysis.

Table 2: Comparison of Stack Tracer with Existing Tools.

Metric Stack Tracer StegExpose DeepStegDetect

Accuracy (%) 95 88 85

FPR (%) 4 9 12

Processing Time (s) 1.2s 2.5s 2.8s

From the results, Stack Tracer demonstrates supe-

rior performance in terms of detection accuracy and

efﬁciency. The low false positive rate indicates its

ability to distinguish benign ﬁles accurately, minimiz-

ing unnecessary alerts. Additionally, the faster pro-

cessing time highlights the system’s computational

efﬁciency compared to competing tools. This makes

Stack Tracer a more effective and scalable solution

for multimedia security analysis.

6 DISCUSSION

The increasing use of steganography in digital me-

dia presents a signiﬁcant challenge for cybersecurity,

necessitating the development of advanced detection

mechanisms such as Stack Tracer. Traditional inspec-

tion techniques often fail to identify hidden data, mak-

ing it imperative to integrate sophisticated steganal-

ysis methods. By combining statistical, frequency-

based, and machine learning-driven techniques, Stack

Tracer enhances the detection accuracy of concealed

content, thereby strengthening digital security.

Recent studies, such as VirusTotal’s (VirusTotal,

2024a) work on browser extension-based malware

detection, emphasize the importance of integrating

automated detection technologies to safeguard users

against embedded threats. Similarly, Norton’s (Nor-

ton, 2022) research on improving safe search mech-

anisms highlights the need for proactive measures to

detect and neutralize concealed risks in multimedia.

These insights underscore the relevance of tools like

Stack Tracer, which address the growing need for ef-

ﬁcient and scalable steganalysis solutions to counter

cyber threats.

7 CONCLUSION AND FUTURE

DEVELOPMENT

The proposed Stack Tracer system presents a ro-

bust approach to detecting hidden ﬁles and assessing

their security risks. By leveraging advanced steganal-

ysis techniques alongside VirusTotal integration, the

system effectively uncovers concealed data in images,

audio, video, and other multimedia formats. The re-

sults demonstrate high detection accuracy with mini-

mal false positives, making Stack Tracer a promising

solution for digital security applications.

To further enhance its capabilities, several key ar-

eas for future development have been identiﬁed:

• Advanced Steganalysis Techniques: Incorporat-

ing additional machine learning-based models and

deep learning frameworks could improve detec-

tion accuracy for more sophisticated and evolving

steganographic methods.

Stack Tracer: Dual-Phase Steganalysis and Malware Detection for Multimedia Security

895

• Real-Time Threat Intelligence: Implementing

real-time updates from VirusTotal and other se-

curity databases will enable proactive detection of

emerging threats and malicious payloads.

• User Interface Enhancements: Reﬁning the in-

terface with improved visualization and real-time

feedback mechanisms will enhance user experi-

ence and threat interpretation.

• Performance and Scalability Optimization: As

the volume and complexity of digital media con-

tinue to rise, optimizing computational efﬁciency

will ensure faster analysis without compromising

accuracy.

• Multi-Layered Security Integration: Extending

compatibility with other cybersecurity tools, such

as endpoint protection solutions and digital foren-

sics platforms, will provide a more holistic secu-

rity framework.

• Extensive User Testing and Feedback Integra-

tion: Conducting real-world testing with cyberse-

curity experts and users will help reﬁne detection

algorithms and usability features, ensuring practi-

cal deployment effectiveness.

By addressing these future enhancements, Stack

Tracer has the potential to become a leading solution

in multimedia security, bridging the gap between ste-

ganalysis and real-time threat detection.

REFERENCES

Abdelfattah, E. and Mahmood, A. (2021). Steganography

and steganalysis: Current status and future directions.

ACM Computing Surveys, 53(4).

Arora, D. N. (2021). Types and tools of steganography. In

Proceedings of the International Conference on Infor-

mation Security and Privacy.

Browsing, G. S. (2022). Safe browsing. Online. Accessed:

2022.

Chrome, G. (2020). What are extensions? Online. Ac-

cessed: 2020-02-06.

Fridrich, J. (2019). Steganalysis. In Handbook of Digital

Forensics and Investigation, pages 451–470. Elsevier.

Hameed, R. S., Ahmad, A. R. B. H., Taher, M. M., and

Mokri, S. S. (2020). A literature review of various

steganography methods. Journal of Computer Appli-

cations, 112(12):14–22.

Kaspersky (2022a). Scam websites: Prevention and safety.

Online. Accessed: 2022.

Kaspersky (2022b). What is smishing and how to defend

against it. Online. Accessed: 2022.

Kaur, N. and Behal, S. (2020). A survey on various

types of steganography and analysis of hiding tech-

niques. International Journal of Computer Applica-

tions, 123(7):28–37.

Meghanathan, N. and Nayak, L. (2021). Steganalysis al-

gorithms for detecting hidden information in image,

audio, and video cover media. In International Con-

ference on Information Security and Privacy, pages

205–214.

Meg

ıas, D., Mazurczyk, W., and Kuribayashi, M. (2020).

Data hiding and its applications: Digital watermark-

ing and steganography. Journal of Information Secu-

rity, 54:330–347.

Norton (2022). Norton safe search enhanced. Online. Ac-

cessed: 2022.

Opera (2020). Extension apis supported in opera. Online.

Accessed: 2020-03-19.

PhishTank (2022). Join the ﬁght against phishing. Online.

Accessed: 2022.

Shehab, D. A. and Alhaddad, M. J. (2019). Comprehen-

sive survey of multimedia steganalysis: Techniques,

evaluations, and trends in future research. Journal of

Computer Security, 27(2):273–290.

Smith, J. and Johnson, A. (2022). Recent developments

in steganography. Journal of Security Research,

34(1):45–58.

Subramanian, N., Al-Maadeed, S., Elharrouss, O., and

Bouridane, A. (2020). Image steganography: A re-

view of the recent advances. IEEE Access, 8:141234–

141250.

Takao, K., Hiraishi, C., Tanabe, R., Takada, K., Fujita, A.,

Inoue, D., Ga

an, C., van Eeten, M., Yoshioka, K.,

and Matsumoto, T. (2017). Vt-sos: A cost-effective

url warning utilizing virustotal as a second opinion

service. In European Symposium on Research in Com-

puter Security, pages 370–388. Springer.

Turner, A. and Lee, M. (2022). A comprehensive review of

malware detection techniques. Cybersecurity Review,

29(2):112–130.

VirusTotal (2024a). Browser extensions. Online. Accessed:

2024.

VirusTotal (2024b). Virustotal. Online. Accessed: 2024.

White, T. and Green, L. (2022). Innovations in multime-

dia steganalysis. International Journal of Information

Security, 43(3):78–92.

INCOFT 2025 - International Conference on Futuristic Technology

896