2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

Object Detection in Images and Videos Using

2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON) | 979-8-3503-8247-1/23/$31.00 ©2023 IEEE | DOI: 10.1109/UPCON59197.2023.10434536

OpenCV: A Comparative Study of Deep

Learning and Traditional Computer Vision
Techniques
Divya Rajawat Bhanu Prakash Lohani Ajay Rana
Undergraduate Scholar Department. of Computer Science Director General
Department of Computer Science Amity University Amity University,
Amity University Greater Noida, India Greater Noida, India
Greater Noida, India [email protected] [email protected]
[email protected]
Prabhat Yadav Shubhi Gupta
Arihant Srivastava Undergraduate Scholor Deptt. of CSE
Undergraduate Scholor Department. of Computer Science Amity University Uttar Pradesh
Department. of Computer Science Amity University Greater Noida
Amity University Greater Noida, India [email protected]
Greater Noida, India [email protected]
[email protected]

Abstract— Using OpenCV, this research compares the Object detection models using CNNs, such as Faster R-
performance of deep learning with standard computer vision CNN, YOLO, and SSD, have obtained outstanding
approaches for detecting object in photos and videos. performance on these datasets. Considering the success of
CNNs, classical computer vision techniques remain popular
Recognizing and localizing objects inside an image or video due to their ease of application and effectiveness [2]. These
is a fundamental task in computer vision. Convolutional neural techniques are often based on handmade characteristics and
networks, for example, have recently demonstrated greater categorize objects in photos using machine learning
accuracy in object detection trials. Traditional computer vision
algorithm like Random Forests , Support Vector Machine
approaches, such as the Viola-Jones algorithm, remain popular
due to their ease of use and performance. In this work, we use
and. The Viola-Jones technique, for example, detects faces
OpenCV, a famous computer vision toolkit, to compare the using Haar-like characteristics and a cascade classifier.
efficacy of both of these methods. On diverse data sets, we OpenCV is a famous open-source computer vision package
evaluate the precision, speed, and complexity of each strategy which offers a vast category of picture and motion graphic
and provide insights into the strengths and drawbacks of each processing operations and tools [7].
method. This paper offers a complete overview of current
object identification strategies and can assist researchers and
practitioners in selecting the best effective approach for their
individual application.

Keywords— OpenCV, CNN, Object Detection, YOLO,

imagenet, SSD

I. INTRODUCTION
Object detection constitutes one of the most difficult and
important jobs in computer vision. It entails locating and
recognizing things of interest within an image or video feed.
There are numerous uses for object detection, including self-
driving cars, monitoring, image search, and robots.
Traditional computer vision approaches have been used to
handle this problem for decades, but the introduction of deep
learning has resulted in a dramatic shift towards using neural
networks to conduct object detection[1]. Fig. 1. A model detecting various objects inside a video frame
CNNs (Convolutional Neural Networks) are the most
widely used deep learning approach for object detection. In this study, we use OpenCV to compare the efficacy of
CNNs are trained to detect objects by using enormous deep learning with classical computer vision approaches for
collections of labelled images. They have demonstrated object detection in photos and videos. Our main goal is to
outstanding performance on a variety of object detection give a detailed study of the most recent object identification
benchmarks, including COCO, PASCAL VOC, and techniques and to explain the strengths and limits of each
ImageNet. methodology. We assess each technique's speed, precision,
and intricacy on a variety of data sets, including COCO,

979-8-3503-8247-1/23/$31.00 ©2023 IEEE 141

Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.
2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

PASCAL VOC, and bespoke data sets, while offering

insights into the trade-offs among various methods.
II. LITERATURE REVIEW
Object detection is an extensively researched subject in
computer vision, with numerous algorithms devised to
handle it. Deep learning-based techniques have recently
attained breakthrough results on a variety of object detection
tests. Traditional computer vision techniques, on the other
hand, are still commonly employed due to their ease of use
and effectiveness [5].
Convolutional neural networks are commonly used in
deep learning-based techniques to object detection. (CNNs).
Convolutional neural networks are sophisticated models
capable of learning complicated features from raw input data.
Backpropagation and stochastic gradient descent are used to Fig. 3. YOLO Object Detection
train such models on huge amounts of labelled images.
Faster R-CNN, YOLO (You Only Look Once), and SSD Khan et al. (2019) conducted another comparative study
(Single Shot Detector) are some prominent deep learning- on the COCO dataset, comparing the accuracy of YOLO
based object identification models [3]. with Faster R-CNN. They discovered that while YOLO was
more rapid, it was less accurate than Faster R-CNN [4]. The
authors additionally evaluated the efficacy of these models to
that of classic computer vision techniques, such as the Viola-
Jones algorithm, and discovered that methods based on deep
learning outperformed traditional computer vision techniques
considerably.

Fig. 2. CNN Architecture

To detect objects in photos and videos, traditional

computer vision systems rely on constructed features and
algorithms. Preprocessing, feature extraction, object
detection, and post- processing processes are often included
in these techniques. Hand-crafted features and machine
learning algorithm like Random Forests & SVMs are
frequently applied in conventional computer vision
techniques for object detection. These techniques are more
comprehensible than deep learning-based approaches in
general, making them appealing in particular situations. The
Viola-Jones algorithm, which detects objects in images using
Haar-like characteristics and a cascade classifier, are
extensively used conventional computer vision methods for
detecting and identifying objects [10].
For object detection within photos and videos, numerous
research have contrasted the performance of classical
computer vision techniques and deep learning-based Fig. 4. Network Diagram of Faster R-CNN
algorithms. Overall, deep learning-based algorithms
outperform classical computer vision techniques, particularly In conclusion, while deep learning-based systems have
for complex objects and scenes. Traditional computer vision achieved state-of-the-art working on a variety of object
techniques, on the other hand, continue to be beneficial in detection benchmarks, classical computer vision techniques
situations where computing performance is a significant remain popular due to their simplicity and speed. Deep
priority, such as real-time video processing [6]. learning-based systems are often more accurate but slower
than classic computer vision algorithms, according to
Several comparison studies on the performance of deep
comparative research.
learning and classical computer vision algorithms for object
detection have been undertaken. Liu et al. (2018), for III. METHODOLOGY
example, compared the efficacy of YOLO and Faster R-
CNN with classic computer vision approaches on the In this study, we use OpenCV to determine the efficacy
PASCAL VOC dataset [9]. of deep learning with classical computer vision approaches
for object detection in photos and videos. To evaluate the
efficacy of alternative approaches, we employ numerous

142
Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.
2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

standard datasets, including COCO and PASCAL VOC, as 4) AR Across Scales:

well as unique datasets. We begin by gathering and preparing (ARS ARM ARL) As AR had variations based on the
datasets for training and testing models. We leverage pre- number of detections per image, here detections are
trained deep learning algorithms like Faster R- CNN and evaluated based on areas same as APAcross Scales.
YOLO that have been built on massive datasets such as
ImageNet. These metrics give us an idea of how the various methods
stack up in comparison to each other for accurate and
To increase their accuracy, we refine these models on the efficient object detection.
unique dataset we are using. For conventional methods of
computer vision, we rely on OpenCV tools and functions to B. Experiment Model Specifications
gather characteristics from images and train algorithms that 1) Faster R-CNN:
use machine learning for classification, such as SVMs and Instead of the standard ROI pooling, crop_and_resize, a
Random Forests. We test various feature extraction Tensorflow function is used. All convolutional layers have
algorithms, such as HOG and SURF, to see which one batch normalization. SGD optimizer is used. The momentum
performs the best. in this scenario is 0.9. The learning rate of this model is
To assess the ability of the different models, we employ a determined by the Feature Extractor.
number of evaluation metrics such as precision, recall, F1 2) R-FCN:
score, and mean average precision. (mAP). We also examine Parameters for this model are similar to the R-CNN.
the rapidity and complexity of each technique to determine Batch normalization is used, so is crop_and_resize. It also
their usefulness in real-world applications. We employ a has the SGD optimizer which has momentum of 0.9. Resnet,
high-performance computer setup with a GPU to train the Inception Resnet and MobileNet feature extractors were used
deep learning models for the trials. We also take advantage for R-FCN.
of OpenCV's parallel processing capabilities to accelerate
classic computer vision approaches. We run each experiment 3) SSD:
numerous times to guarantee that the results are statistically All layers make use of Batch Normalization. In the
significant. We do a full analysis of the data, including beginning, the weights are set as 0.3 (standard deviation).
visualizations of the discovered items and their bounding The feature maps are convolutional. They are implemented
boxes, to compare the performance of the different for prediction. The decay factor is set to 2.
techniques. We also do a sensitivity analysis to assess the
effect of various factors and configurations on the Average Precision Values
performance of each technique. 70
To train the deep learning models for the trials, we use a 60
high- performance computer configuration with a GPU. We 50
also use OpenCV's parallel processing capabilities to speed 40
up traditional computer vision methodologies. Each
experiment is repeated several times to ensure that the results 30
are statistically significant. To assess the efficacy of each Faster R- Faster R- Faster R- Faster R-
method, we do an in-depth examination of the data, CNN with CNN with CNN by G- CNN with ResNet-
including representations of the found items and their AP@50:5:95 AP50 AP75
bounding boxes. We also run an impact study to determine AP-S AP-M AP-L
how different factors and configurations affect the efficacy
of each technique.
A. Open-source Performance metrics of various Average Precision Values
algorithms
70
Bounding Box, or The COCO detection challenge, is a
competition where people can attempt to detect objects in 60
over 2,00,000 images comprising 80 different categories. 50
The bounding box coordinates are available for those 40
images. Works submitted ranked according to four major 30
metrics groups.
20
1) AP: 10
The IOU range varies from 50% to 95% with steps of
5%, also termed as AP@50:5:95. Single values of IOU can Faster R- Faster R- Faster R- Faster R-
CNN with CNN with CNN by G- CNN with
also be evaluated. Fifty percent and Seventy-five percent are
ResNet-101 FP RMI TDM
used frequently. They are represented by AP50 and AP75
AP@50:5:95 AP50 AP75
2) AP Across Scales:
The Average Precision is calculated according to object AP-S AP-M AP-L
size: less than 322 pixels for small, between 322 and 962
pixels for medium and more than 962 pixels for large. Fig. 5. Average Precision values for different methods

3) Average Recall (AR):

Based on the amount of detections per image, recall
values are calculated. The IOUs used are the same as AP.

143
Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.
2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

traditional vision algorithms, only minimal data is required.

Accuracy and mAP These models can be seen as the quick and dirty way of
doing computer vision, though they are surprisingly
on COCO Dataset effective. They are just limited on accuracy and range of
objects that they can detect. Deep learning algorithms like
MobileNet CNN on the other hand, are very accurate. However, they
require large amount of data in order to train effectively and
Inception Resnet give desired results.
Resnet-101 In case of different deep learning models; on the COCO
dataset, various combinations of them were tested. We found
VGG-16 that SSD outperforms the most accurate models when using
lightweight feature extractors on larger images. Moreover,
0 10 20 30 40 we discovered that speed can be increased with fewer
Imagenet Accuracy Min mAP Score(COCO)
proposals while maintaining same mAP scores. We want to
experiment with various pairings to discover better outcomes
Max mAP Score(COCO) and sweet spots that may be used for particular use cases.

Fig. 6. Max and min mAP scores on COCO dataset of different extractor REFERENCES
modelt [1] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image
recognition,” in IEEE Conference on Computer Vision and Pattern
SSD MobileNet was the fastest; Faster R-CNN Recognition, Jun 2016, pp. 770–778
Inception Resnet was the slowest. First one was trained [2] S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-
with resolution of 300 whereas the second was trained with time object detection with region proposal networks,” in Advances in
Neural Information Processing Systems 28, 2015, pp. 91–99.
600. Lowest accuracy was achieved by R-FCN MobileNet,
[3] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. E. Reed, C. Fu, and
and Faster R- CNN Inception Resnet had the highest A. C. Berg, “SSD: single shot multibox detector,” CoRR, 2015.
accuracy. [4] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You only look
Faster R-CNN gives out the most accurate results. It once: Unified, real-time object detection,” in IEEE Conference on
Computer Vision and Pattern Recognition, 2016, pp. 779–788.
also takes a lot of time to do so, about 100 milliseconds per
[5] Redmon, J. (n.d.). YOLO: Real-Time Object Detection.
image. In contrast, SSD and R-FCN models are usually http://pjreddie.com/yolo/
faster which might come at the cost of accuracy of the [6] Wu W, Dasgupta D, Ramirez E, et al. Classification accuracies of
model. The slower more accurate models can be tweaked physical activities using smartphone motion sensors[J]. Journal of
to take less time by varying the number of proposed medical Internet research,2012,14(5): el30-el30.
regions. The correlation of accuracy of feature extractors [7] GeeksforGeeks. (2023). Detect an object with OpenCV Python.
and the mean Average Precision scores on COCO dataset GeeksforGeeks. https://www.geeksforgeeks.org/detect-an-object-
stands true only for the models that do not have a greater with-opencv-python/
dependency on the classification from their feature [8] Kumar, V. (2022, March 30). How to Detect Objects in Real-Time
extractors. Using OpenCV and Python. Medium.
https://towardsdatascience.com/how-to-detect-objects-in-real-time-
using-opencv-and-python-c1ba0c2c69c0.
GPU Time and mAP Scores [9] Great Learning Team. (2022, March 22). Object Detection Using
OpenCV YOLO | Great Learning. Great Learning Blog: Free
1000 Resources What Matters to Shape Your Career!
https://www.mygreatlearning.com/blog/yolo-object-detection-using-
900 opencv/
800 [10] Waseem, M. (2023). Object Detection with OpenCV-Python Using a
Haar-Cascade Classifier. Stack Abuse. https://stackabuse.com/object-
700 detection-with-opencv-python-using-a-haar-cascade-classifier/
[11] P. K. Kushwaha and M. Kumaresan, "Machine learning algorithm in
600 healthcare system: A Review," 2021 International Conference on
Technological Advancements and Innovations (ICTAI), Tashkent,
500 Uzbekistan, 2021, pp. 478-481, doi:
400 10.1109/ICTAI53825.2021.9673220.
[12] P. K. Kushwaha, B. P. Lohani and D. Singh, "Review on information
300 security, laws and ethical issues with online financial system," 2016
International Conference on Innovation and Challenges in Cyber
200 Security (ICICCS-INBUSH), Greater Noida, India, 2016, pp. 49-53,
0 2 4 6 8 10 12 doi: 10.1109/ICICCS.2016.7542350.
[13] G. Gulati, B. P. Lohani and P. K. Kushwaha, “A Novel Application
mAP Score GPU Time Of IoT In Empowering Women Safety Using GPS Tracking Module,”
2020 Research, Innovation, Knowledge Management and Technology
Fig. 7. GPU times and mean Average Precision values for all the Application for Business Sustainability (INBUSH), Greater Noida,
combinations India, 2020, pp. 131-137, doi:
10.1109/INBUSH46973.2020.9392193.
IV. CONCLUSION [14] D. Pareta, I. N. Verma, B. P. Lohani, P. K. Kushwaha and V. Bibhu,
"IoT Enabled Smart and Efficient Musical Water Fountain," 2022 2nd
There exists a trade-off between accuracy and International Conference on Innovative Practices in Technology and
investment of time and resources in the case of traditional Management (ICIPTM), Gautam Buddha Nagar, India, 2022, pp. 369-
computer vision algorithms versus CNNs. In order to train 373, doi: 10.1109/ICIPTM54933.2022.9754129.

144
Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.
2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

[15] B. P. Lohani, M. Trivedi, R. J. Singh, V. Bibhu, S. Ranjan and P. K. Analysis and Knowledge Management (ABLAZE), Greater Noida,
Kushwaha, "Machine Learning Based Model for Prediction of Loan India, 2015, pp. 206-211, doi: 10.1109/ABLAZE.2015.7154993.
Approval," 2022 3rd International Conference on Intelligent [31] Behl, R., Khatter, H., Singh, P., Bhardwaj, G., Chaturvedi, P. (2021).
Engineering and Management (ICIEM), London, United Kingdom, Automated and Curated Sack Count Leveraging Video Analysis on
2022, pp. 465-470, doi: 10.1109/ICIEM54221.2022.9853160. Moving Objects. In: Awasthi, S., Travieso-González, C.M., Sanyal,
[16] A. Kumar, B. P. Lohani and P. K. Kushwaha, "Robust Secured G., Kumar Singh, D. (eds) Artificial Intelligence for a Sustainable
Framework for Online Business Transactions over Public Network," Industry 4.0. Springer, Cham. https://doi.org/10.1007/978-3-030-
2021 2nd International Conference on Intelligent Engineering and 77070-9_17.
Management (ICIEM), London, United Kingdom, 2021, pp. 555-560, [32] L. Das, S. Sharma, A. Naval, A. Singh and P. Anand, "Distributive
doi: 10.1109/ICIEM51511.2021.9445380. and Governing System for Descriptive Error Identification of High
[17] P. K. Kushwaha and B. P. Lohani, "A review of security of the cloud Speed Railway Illustrations and Images using Convolutional Neural
computing over business with implementation," 2016 International Networks," 2022 3rd International Conference on Intelligent
Conference on Innovation and Challenges in Cyber Security Engineering and Management (ICIEM), London, United Kingdom,
(ICICCS-INBUSH), Greater Noida, India, 2016, pp. 192-198, doi: 2022, pp. 815-820, doi: 10.1109/ICIEM54221.2022.9853125.
10.1109/ICICCS.2016.7542342. [33] N. Krishnachaithanya et al., "People Counting in Public Spaces using
[18] M. Chandra, P. K. Kushwaha and S. Saxena, "Modified Fractal Deep Learning-based Object Detection and Tracking Techniques,"
Carpets," 2011 International Conference on Computational 2023 International Conference on Computational Intelligence and
Intelligence and Communication Networks, Gwalior, India, 2011, pp. Sustainable Engineering Solutions (CISES), Greater Noida, India,
537-540, doi: 10.1109/CICN.2011.115. 2023, pp. 784-788, doi: 10.1109/CISES58720.2023.10183503.
[19] P. K. Kushwaha, R. Kohli and D. Singh, "Secret key watermarking in [34] K. Kaushik, I. Punhani, S. Sharma and M. Martolia, "An Advanced
WAV audio file in perceptual domain," 2015 International Approach for performing Cyber Fraud using Banner Grabbing," 2022
Conference on Futuristic Trends on Computational Analysis and 5th International Conference on Contemporary Computing and
Knowledge Management (ABLAZE), Greater Noida, India, 2015, pp. Informatics (IC3I), Uttar Pradesh, India, 2022, pp. 298-302, doi:
629-634, doi: 10.1109/ABLAZE.2015.7154940. 10.1109/IC3I56241.2022.10072445.
[20] Ranjan, Ankur A. et al. “An Approach for Netflix Recommendation [35] S. Manna, V. Jalodia, K. Kumar, V. Tripathi, S. Sharma and D.
System using Singular Value Decomposition.” Journal of Computer Arora, "Predicting preeminent Machine Learning Approach on Stars,"
and Mathematical Sciences (2019). 2022 2nd International Conference on Technological Advancements
[21] Makkar, Bhavya et al. “Map Reduce concept-based Sentiment in Computational Sciences (ICTACS), Tashkent, Uzbekistan, 2022,
Analysis Approach.” International Journal of Computer Sciences and pp. 587-591, doi: 10.1109/ICTACS56270.2022.9988044.
Engineering (2019) [36] A. Chaturvedi, S. A. Yadav, H. M. Salman, H. R. Goyal, H.
[22] Bhatia, Ayush & Bibhu, Vimal & Lohani, Bhanu & Kushwaha, Gebregziabher and A. K. Rao, "Classification of Sound using
Pradeep. (2020). An Application Framework for Quantum Computing Convolutional Neural Networks," 2022 5th International Conference
using Artificial intelligence Techniques. 264-269. on Contemporary Computing and Informatics (IC3I), Uttar Pradesh,
10.1109/INBUSH46973.2020.9392164. India, 2022, pp. 1015-1019, doi: 10.1109/IC3I56241.2022.10072823.
[23] A. Kumar, B. P. Lohani and P. K. Kushwaha, "Black Hole Attack in [37] V. Agarwal, S. Taware, S. A. Yadav, D. Gangodkar, A. Rao and V.
Mobile Ad Hoc Network and its Avoidance," 2021 International K. Srivastav, "Customer - Churn Prediction Using Machine
Conference on Innovative Practices in Technology and Management Learning," 2022 2nd International Conference on Technological
(ICIPTM), Noida, India, 2021, pp. 103-107, doi: Advancements in Computational Sciences (ICTACS), Tashkent,
10.1109/ICIPTM52218.2021.9388366. Uzbekistan, 2022, pp. 893-899, doi:
10.1109/ICTACS56270.2022.9988187.
[24] Srivastav, A.V., Lohani, B.P., Kushwaha, P.K., Tyagi, S. (2021).
Dual-Layer Security and Access System to Prevent the Spread of [38] A. Shahi, S. Kaur, A. Mittal and S. V. Singh, "Building Technology
COVID-19. In: Prateek, M., Singh, T.P., Choudhury, T., Pandey, adoption model for the success of Women Healthcare Workers," 2022
H.M., Gia Nhu, N. (eds) Proceedings of International Conference on 5th International Conference on Contemporary Computing and
Machine Intelligence and Data Science Applications. Algorithms for Informatics (IC3I), Uttar Pradesh, India, 2022, pp. 175-180, doi:
Intelligent Systems. Springer, Singapore. 10.1109/IC3I56241.2022.10073124.
https://doi.org/10.1007/978-981-33-4087-9_28 [39] V. Malik, R. Mittal and S. V. SIngh, "EPR-ML: E-Commerce Product
[25] A. Khuran, B. P. Lohani, V. Bibhu and P. K. Kushwaha, "An AI Recommendation Using NLP and Machine Learning Algorithm,"
Integrated Face Detection System for Biometric Attendance 2022 5th International Conference on Contemporary Computing and
Management," 2021 2nd International Conference on Intelligent Informatics (IC3I), Uttar Pradesh, India, 2022, pp. 1778-1783, doi:
Engineering and Management (ICIEM), London, United Kingdom, 10.1109/IC3I56241.2022.10073224.
2021, pp. 29-33, doi: 10.1109/ICIEM51511.2021.9445295. [40] R. Mittal, V. Malik and S. V. Singh, "DFR-HL: Diabetic Food
[26] S. Salagrama, B. P. Lohani and P. K. Kushwaha, "An Analytical Recommendation Using Hybrid Learning Methods," 2022 5th
Survey of User Privacy on Social Media Platform," 2021 International Conference on Contemporary Computing and
International Conference on Technological Advancements and Informatics (IC3I), Uttar Pradesh, India, 2022, pp. 1784-1788, doi:
Innovations (ICTAI), Tashkent, Uzbekistan, 2021, pp. 173-176, doi: 10.1109/IC3I56241.2022.10072763.
10.1109/ICTAI53825.2021.9673402. [41] R. Mittal, J. Singh, V. Malik, A. Mittal, V. Rattan and S. V. Singh,
[27] S. Singh, D. Chaudhary, A. D. Gupta, B. Prakash Lohani, P. K. "Forecasting E-Mentoring Effectiveness using Data Mining
Kushwaha and V. Bibhu, "Artificial Intelligence, Cognitive Robotics Approach," 2022 5th International Conference on Contemporary
and Nature of Consciousness," 2022 3rd International Conference on Computing and Informatics (IC3I), Uttar Pradesh, India, 2022, pp.
Intelligent Engineering and Management (ICIEM), London, United 931-934, doi: 10.1109/IC3I56241.2022.10072806.
Kingdom, 2022, pp. 447-454, doi: [42] D. Jain, M. Arya, V. Malik and S. V. Singh, "A Novel Parameter
10.1109/ICIEM54221.2022.9853081. Optimization Metaheuristic: Human Habitation Behavior Based
[28] S. Suman, P. Kaushik, S. S. N. Challapalli, B. P. Lohani, P. Optimization," 2022 5th International Conference on Contemporary
Kushwaha and A. D. Gupta, "Commodity Price Prediction for making Computing and Informatics (IC3I), Uttar Pradesh, India, 2022, pp.
informed Decisions while trading using Long Short-Term Memory 921-924, doi: 10.1109/IC3I56241.2022.10072699.
(LSTM) Algorithm," 2022 5th International Conference on [43] G. Singh, G. Bhardwaj, S. V. Singh and N. Chaudhary, "Artificial
Contemporary Computing and Informatics (IC3I), Uttar Pradesh, Intelligence led Industry 4.0 Application for Sustainable
India, 2022, pp. 406-411, doi: 10.1109/IC3I56241.2022.10072626. Development," 2022 2nd International Conference on Innovative
[29] Gupta, A., Gupta, M., & Chaturvedi, P. (2020). Investing Data with Practices in Technology and Management (ICIPTM), Gautam Buddha
Machine Learning Using Python. Strategic System Assurance and Nagar, India, 2022, pp. 339-343, doi:
Business Analytics, 1-9. 10.1109/ICIPTM54933.2022.9753944.
[30] P. Chaturvedi, S. Dahiya and S. Agrawal, "Technological innovation: [44] T. Vinoth Kumar, A. R. Yeruva, S. Kumar, D. Gangodkar, A. L N
A necessity for sustainable MSME sector in India," 2015 Rao and P. Chaturvedi, "A New Vehicle Tracking System with R-
International Conference on Futuristic Trends on Computational CNN and Random Forest Classifier for Disaster Management
Platform to Improve Performance," 2022 2nd International

145
Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.
2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON)

Conference on Technological Advancements in Computational in Computational Sciences (ICTACS), Tashkent, Uzbekistan, 2022,
Sciences (ICTACS), Tashkent, Uzbekistan, 2022, pp. 797-804, doi: pp. 658-662, doi: 10.1109/ICTACS56270.2022.9988033.
10.1109/ICTACS56270.2022.9988024. [49] S. H. Abbas, S. Vashisht, G. Bhardwaj, R. Rawat, A. Shrivastava and
[45] A. R. Yeruva, C. S. L Vijaya Durga, G. B, K. Pant, P. Chaturvedi and K. Rani, "An Advanced Cloud-Based Plant Health Detection System
A. P. Srivastava, "A Smart Healthcare Monitoring System Based on Based on Deep Learning," 2022 5th International Conference on
Fog Computing Architecture," 2022 2nd International Conference on Contemporary Computing and Informatics (IC3I), Uttar Pradesh,
Technological Advancements in Computational Sciences (ICTACS), India, 2022, pp. 1357-1362, doi: 10.1109/IC3I56241.2022.10072786.
Tashkent, Uzbekistan, 2022, pp. 904-909, doi: [50] R. Singh, M. Bansal, S. Gupta, A. Singh, G. Bhardwaj and A. D.
10.1109/ICTACS56270.2022.9987881. Dhariwal, "Detection of Social Network Spam Based on Improved
[46] D. Singh, P. Chaturvedi, G. Dhawan, A. Singh and R. Aggarwal, Machine Learning," 2022 5th International Conference on
"WBMS: Waste Bin Management System for densely populated Contemporary Computing and Informatics (IC3I), Uttar Pradesh,
urban areas," 2022 2nd International Conference on Innovative India, 2022, pp. 2257-2261, doi: 10.1109/IC3I56241.2022.10073448.
Practices in Technology and Management (ICIPTM), Gautam Buddha [51] S. Gupta, G. Bhardwaj, A. Shrivastava, R. Kukreti, A. K. Rao and B.
Nagar, India, 2022, pp. 770-774, doi: Chauhan, "Liver Disease Prediction Using Various Algorithms," 2022
10.1109/ICIPTM54933.2022.9754190. 5th International Conference on Contemporary Computing and
[47] S. Agarwal, G. Bhardwaj, E. Saraswat, N. Singh, R. Aggarwal and A. Informatics (IC3I), Uttar Pradesh, India, 2022, pp. 827-831, doi:
Bansal, "Insurtech Fostering Automated Insurance Process using 10.1109/IC3I56241.2022.10072433.
Deep Learning Approach," 2022 2nd International Conference on [52] U. N. Dulhare, S. Rasool, M. N. Khan, B. Pant, A. Kakoli Rao and G.
Innovative Practices in Technology and Management (ICIPTM), Bhardwaj, "Analysis of the Regulatory Development
Gautam Buddha Nagar, India, 2022, pp. 386-392, doi: Cryptocurrencies for Trading in Business with Deep Learning
10.1109/ICIPTM54933.2022.9753891. Techniques," 2022 2nd International Conference on Technological
[48] M. Patidar, G. Bhardwaj, A. Jain, B. Pant, D. Kumar Ray and S. Advancements in Computational Sciences (ICTACS), Tashkent,
Sharma, "An Empirical Study and Simulation Analysis of the MAC Uzbekistan, 2022, pp. 651-657, doi:
Layer Model Using the AWGN Channel on WiMAX Technology," 10.1109/ICTACS56270.2022.9987953.
2022 2nd International Conference on Technological Advancements

146
Authorized licensed use limited to: Universita degli Studi di Napoli Federico II. Downloaded on March 30,2025 at 22:36:11 UTC from IEEE Xplore. Restrictions apply.