A Deep Learning Framework for Multi-Class Objectrecognition Using Feature Fusion and Neigborhoodcomponent Analysis DOI
Anamika Dhillon, Gyanendra K. Verma

SSRN Electronic Journal, Journal Year: 2021, Volume and Issue: unknown

Published: Jan. 1, 2021

Object recognition is a computer vision technique for identifying objects inan image. Deep neural networks have demonstrated remarkable recognitionresults on the basis of features extracted from single image object. In this paper, we present feature fusion-based deep learning method forclassifying and recognizing multi-class objects. Specifically, first adopttwo Convolutional Neural Network models: DenseNet 201 ResNet 101, forfeature extraction. Then, to acquire more compact presentation featuresand reduce dimensions, utilized Neighborhood Component Analysis(NCA). Furthermore, fusion performed in hierarchical manner byapplying concatenation operation. Finally, classify multiple objectsin an by using Support Vector Machine (SVM) classifier. We demonstrate effectiveness our methodology two benchmark datasets; MSCOCO wild animal camera trap dataset. The experimental results showthat proposed framework achieved accuracy 98.1% 97.5% datasets respectively. Results showed thatour effectively improved performance favorably bothrobustness accuracy. fair comparison with existing techniques reported literature also provided

Language: Английский

Advanced Dual Module Weapon Detection for Public Safety and Surveillance System DOI
Manikanta Sirigineedi,

M. Sowmiya,

U. Hemavathi

et al.

Published: Oct. 15, 2024

Language: Английский

Citations

0
Fundamental investigation of hair type classification method using deep learning for human hair screening at crime scenes DOI

Yoshito Tomisaka,

Yoshiki Chushi,

Ami Nagata

et al.

Japanese Journal of Forensic Science and Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 21, 2024

Language: Английский

Citations

0
Automated detection of regions of interest in cartridge case images using deep learning DOI Creative Commons
Marie‐Eve Le Bouthillier, Lynne Hrynkiw, A.L. Beauchamp

et al.

Journal of Forensic Sciences, Journal Year: 2023, Volume and Issue: 68(6), P. 1958 - 1971

Published: July 12, 2023

This paper explores a deep-learning approach to evaluate the position of circular delimiters in cartridge case images. These define two regions interest (ROI), corresponding breech face and firing pin impressions, are placed manually or by an image-processing algorithm. positioning bears significant impact on performance image-matching algorithms for firearm identification, automated evaluation method would be beneficial any computerized system. Our contribution consists optimizing training U-Net segmentation models from digital images cases, intending locate ROIs automatically. For experiments, we used high-resolution 2D 1195 samples cases fired different 9MM firearms. results show that models, trained augmented data sets, exhibit 95.6% IoU (Intersection over Union) 99.3% DC (Dice Coefficient) with loss 0.014 images; 95.9% 99.5% 0.011 We observed natural shapes predicted circles reduce compared perfect ground truth masks suggesting our provide more accurate real ROI shape. In practice, believe these could useful firearms identification. future work, predictions may quality specimens database, they determine region image.

Language: Английский

Citations

1
Analysis of Traces on Discharged Bullets by the Congruent Matching Profile Segments Method and k-Nearest Neighbors DOI
В. А. Федоренко, K. O. Sorokina, Pavel Giverts

et al.

Programming and Computer Software, Journal Year: 2023, Volume and Issue: 49(S2), P. S72 - S81

Published: Dec. 1, 2023

Language: Английский

Citations

0
A Deep Learning Framework for Multi-Class Objectrecognition Using Feature Fusion and Neigborhoodcomponent Analysis DOI
Anamika Dhillon, Gyanendra K. Verma

SSRN Electronic Journal, Journal Year: 2021, Volume and Issue: unknown

Published: Jan. 1, 2021

Object recognition is a computer vision technique for identifying objects inan image. Deep neural networks have demonstrated remarkable recognitionresults on the basis of features extracted from single image object. In this paper, we present feature fusion-based deep learning method forclassifying and recognizing multi-class objects. Specifically, first adopttwo Convolutional Neural Network models: DenseNet 201 ResNet 101, forfeature extraction. Then, to acquire more compact presentation featuresand reduce dimensions, utilized Neighborhood Component Analysis(NCA). Furthermore, fusion performed in hierarchical manner byapplying concatenation operation. Finally, classify multiple objectsin an by using Support Vector Machine (SVM) classifier. We demonstrate effectiveness our methodology two benchmark datasets; MSCOCO wild animal camera trap dataset. The experimental results showthat proposed framework achieved accuracy 98.1% 97.5% datasets respectively. Results showed thatour effectively improved performance favorably bothrobustness accuracy. fair comparison with existing techniques reported literature also provided

Language: Английский

Citations

0