Real Time Multi Object Detection for Blind Using Single Shot Multibox Detector

Arora, Adwitiya; Grover, Atul; Chugh, Raksha; Reka, S. Sofana

doi:10.1007/s11277-019-06294-1

Cited by 48 publications

(13 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The real-time dataset is used for the testing. Adwitiya et al [24] have introduced a prototype for the detection of an object/obstacle in real-time based on the Deep Neural Network. The prototype uses a single shot multi-box framework for the detection purpose.…”

Section: ) Non-vision-based Detection Devicesmentioning

confidence: 99%

Assistive Devices Analysis for Visually Impaired Persons: A Review on Taxonomy

Zafar

Asif²,

Ahmad³

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Visually impaired persons (VIPs) comprise a significant portion of the population and they are present in all corners of the world. In recent times, technology proved its presence in every domain of the life and innovative devices are assisting humans in all fields especially, artificial intelligence has dominated and outperformed rest of the trades. VIPs need assistance in performing daily life tasks like object/obstacle detection and recognition, navigation, and mobility, particularly in indoor and outdoor environments. Moreover, the protection and safety of these people are of prime concern. Several devices and applications have been developed for the assistance of VIPs. Firstly, these devices take input from the surrounding environment through different sensors e.g. infrared radiation, ultrasonic, imagery sensor, etc. In the second stage, state of the art machine learning techniques process these signals and extract useful information. Finally, feedback is provided to the user through auditory and/or vibratory means. It is observed that most of the existing devices are constrained in their abilities. The paper presents a comprehensive comparative analysis of the state-of-the-art assistive devices for VIPs. These techniques are categorized based on their functionality and working principles. The main attributes, challenges, and limitations of these techniques have also been highlighted. Moreover, a score based quantitative analysis of these devices is performed to highlight their feature enrichment capability for each category. It may help to select an appropriate device for a particular scenario.

show abstract

Section: ) Non-vision-based Detection Devicesmentioning

confidence: 99%

Assistive Devices Analysis for Visually Impaired Persons: A Review on Taxonomy

Zafar

Asif²,

Ahmad³

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Using multibox [43], the SSD takes only one shot to detect multiple objects present in an image. e SSD has a substantially faster object detection algorithm with high accuracy.…”

Section: Ssd-rnn For Obstacle Recognitionmentioning

confidence: 99%

Vision Navigator: A Smart and Intelligent Obstacle Recognition Model for Visually Impaired Users

Suman

Mishra

Sahoo

et al. 2022

Mobile Information Systems

View full text Add to dashboard Cite

Vision impairment is a major challenge faced by humanity on a large scale throughout the world. Affected people find independently navigating and detecting obstacles extremely tedious. Thus, a potential solution for accurately detecting obstacles requires an integrated deployment of the Internet of Things and predictive analytics. This research introduces “Vision Navigator,” a novel framework for assisting visually impaired users in obstacle analysis and tracking so that they can move independently. An intelligent stick named “Smart-fold Cane” and sensor-equipped shoes called “Smart-alert Walker” are the main constituents of our proposed model. For object detection and classification, the stick uses a single-shot detection (SSD) mechanism, which is followed by frame generation using the recurrent neural network (RNN) model. Smart-alert Walker is a lightweight shoe that acts as an emergency unit that notifies the user regarding the presence of any obstacle within a short distance range. This intelligent obstacle detection model using the SSD-RNN approach was deployed in real time and its performance was validated in indoor and outdoor environments. The SSD-RNN model computed an optimum accuracy of 95.06% and 87.68% indoors and outdoors, respectively. The model was also evaluated in the context of users’ distance from obstacles. The proposed SSD-RNN model had an accuracy rate of 96.4% and 86.8% for close and distant obstacles, respectively, outperforming other models. Execution time for the SSD-RNN model was 4.82 s with the highest mean accuracy rate of 95.54% considering all common obstacles.

show abstract

“…Figure 5: MobileNet-SSD Architecture SSDis used for object detection and classification together [9]. We are an interested predictor to detect all objects even faster than MobileNet that's why we combined SSD and MobileNet has satisfactorily predicted the bounding boxes ofthe objects [10]. MobileNet as backbone architecture for SSD and achieve comparable detection accuracy on our data-set and in this paper, MobileNet is the base network of SSD [7].It can also be extended as a competent mean Reticulum in recent object detection systems [11] With small anchor networks, lightweight that can be utilized for categorization, perception and separation tasks and depth-wise dissociable convolution layer as used MobileNet-SSD, also one-stage detectors [10,12].…”

Section: Mask R-cnnmentioning

confidence: 99%

“…The SSD is a regression-based algorithm. The network generates the collection of bounding boxes with three different aspect ratios and confidence score for each bounding boxes, which are used to filter the anchor boxes with object class instances, and then a non-maximum suppression step to generate the final detection [10]. Single shot means that in the recognition process, the localization and classification of the object are carried out in one pass, and the network only makes one "look" of the image.…”

Section: Our Workmentioning

confidence: 99%

Electric Equipment Inspection on High Voltage Transmission Line Via Mobile Net-SSD

Zhan¹

2021

converter

View full text Add to dashboard Cite

Daily check and inspection of electrical utilities on the transmission line to find out faults or malfunction data and analyze, it’s to ensure normal state of electrical equipment really difficult in any situation. Machine-controlled inspections by like robots or drones for power transmission infrastructures is an indispensable way to assure the safety of power transmission. Targeted object detection and classification of the power transmission infrastructure is the prerequisite for automatic inspection. In our experiment we have create the dedicated datasets of the electric equipment on power transmission line for multi-object detection, including our data collection, prepossessing and annotation. This work has been done multiple experiments to solve our functional problem and compare novel state of art deep learning methods such as Faster R-CNN, Mask R-CNN, YOLO, and SSD with MobileNet is a base feature extractor, to realize the electric equipment on power transmission line detection. For Condition monitoringand diagnosis identification of the importance of electric equipment on the electric transfer line, in the proposed deep detection approach, the Single-Shot Multi-box Detector (SSD) is a powerful deepmeta-architecture. The results show that our method can automatically detect electric equipment on high voltage transfer defects more accurately and rapidly than lightweight network methods and traditional deep learning methods. Results shed new light on defect detection in actual in progressive scenarios. In our research the main goal to show the implementation of the object detection on electric equipment's inspections on high voltage electric transfer lines on drone video using MobileNet-SSD object detection and recognition.

show abstract

Real Time Multi Object Detection for Blind Using Single Shot Multibox Detector

Cited by 48 publications

References 27 publications

Assistive Devices Analysis for Visually Impaired Persons: A Review on Taxonomy

Assistive Devices Analysis for Visually Impaired Persons: A Review on Taxonomy

Vision Navigator: A Smart and Intelligent Obstacle Recognition Model for Visually Impaired Users

Electric Equipment Inspection on High Voltage Transmission Line Via Mobile Net-SSD

Contact Info

Product

Resources

About