Simplifying and Streamlining Large-Scale Materials Image Processing with Wizard-Driven and Scalable Deep Learning

Automated driving and vehicle safety systems need object detection. It is important that object detection be accurate overall and robust to weather and environmental conditions and run in real-time. As a consequence of this approach, they require image processing algorithms to inspect the contents of images. This article compares the accuracy of five major image processing algorithms: Region-based Fully Convolutional Network (R-FCN), Mask Region-based Convolutional Neural Networks (Mask R-CNN), Single Shot Multi-Box Detector (SSD), RetinaNet, and You Only Look Once v4 (YOLOv4). In this comparative analysis, we used a large-scale Berkeley Deep Drive (BDD100K) dataset. Their strengths and limitations are analyzed based on parameters such as accuracy (with/without occlusion and truncation), computation time, precision-recall curve. The comparison is given in this article helpful in understanding the pros and cons of standard deep learning-based algorithms while operating under real-time deployment restrictions. We conclude that the YOLOv4 outperforms accurately in detecting difficult road target objects under complex road scenarios and weather conditions in an identical testing environment.

Download Full-text

DiPLIP: Distributed Parallel Processing Platform for Stream Image Processing Based on Deep Learning Model Inference

Electronics ◽

10.3390/electronics9101664 ◽

2020 ◽

Vol 9 (10) ◽

pp. 1664

Author(s):

Yoon-Ki Kim ◽

Yongsung Kim

Keyword(s):

Image Processing ◽

Deep Learning ◽

Parallel Processing ◽

Real Time ◽

Large Scale ◽

Learning Model ◽

Streaming Data ◽

Model Inference ◽

Processing Platform ◽

Deep Learning Model

Recently, as the amount of real-time video streaming data has increased, distributed parallel processing systems have rapidly evolved to process large-scale data. In addition, with an increase in the scale of computing resources constituting the distributed parallel processing system, the orchestration of technology has become crucial for proper management of computing resources, in terms of allocating computing resources, setting up a programming environment, and deploying user applications. In this paper, we present a new distributed parallel processing platform for real-time large-scale image processing based on deep learning model inference, called DiPLIP. It provides a scheme for large-scale real-time image inference using buffer layer and a scalable parallel processing environment according to the size of the stream image. It allows users to easily process trained deep learning models for processing real-time images in a distributed parallel processing environment at high speeds, through the distribution of the virtual machine container.

Download Full-text

Location Invariant Animal Recognition Using Mixed Source Datasets and Deep Learning

10.1101/2020.05.13.094896 ◽

2020 ◽

Author(s):

Andrew Shepley ◽

Greg Falzon ◽

Paul Meek ◽

Paul Kwan

Keyword(s):

Image Processing ◽

Deep Learning ◽

Object Detection ◽

Transfer Learning ◽

Large Scale ◽

Camera Trap ◽

List Type ◽

Learning Models ◽

Geographical Regions ◽

Flickr Images

AbstractA time-consuming challenge faced by camera trap practitioners all over the world is the extraction of meaningful data from images to inform ecological management. The primary methods of image processing used by practitioners includes manual analysis and citizen science. An increasingly popular alternative is automated image classification software. However, most automated solutions are not sufficiently robust to be deployed on a large scale. Key challenges include limited access to images for each species and lack of location invariance when transferring models between sites. This prevents optimal use of ecological data and results in significant expenditure of time and resources to annotate and retrain deep learning models.In this study, we aimed to (a) assess the value of publicly available non-iconic FlickR images in the training of deep learning models for camera trap object detection, (b) develop an out-of-the-box location invariant automated camera trap image processing solution for ecologist using deep transfer learning and (c) explore the use of small subsets of camera trap images in optimisation of a FlickR trained deep learning model for high precision ecological object detection.We collected and annotated a dataset of images of “pigs” (Sus scrofa and Phacochoerus africanus) from the consumer image sharing website FlickR. These images were used to achieve transfer learning using a RetinaNet model in the task of object detection. We compared the performance of this model to the performance of models trained on combinations of camera trap images obtained from five different projects, each characterised by 5 different geographical regions. Furthermore, we explored optimisation of the FlickR model via infusion of small subsets of camera trap images to increase robustness in difficult images.In most cases, the mean Average Precision (mAP) of the FlickR trained model when tested on out of sample camera trap sites (67.21-91.92%) was significantly higher than the mAP achieved by models trained on only one geographical location (4.42-90.8%) and rivalled the mAP of models trained on mixed camera trap datasets (68.96-92.75%). The infusion of camera trap images into the FlickR training further improved AP by 5.10-22.32% to 83.60-97.02%.Ecology researchers can use FlickR images in the training of automated deep learning solutions for camera trap image processing to significantly reduce time and resource expenditure by allowing the development of location invariant, highly robust out-of-the-box solutions. This would allow AI technologies to be deployed on a large scale in ecological applications.

Download Full-text

Deep Learning based Image Processing for Cashier-less Self-Checkout Methodology

International Journal of Recent Technology and Engineering - 2 ◽

10.35940/ijrte.a2958.059120 ◽

2020 ◽

Vol 9 (1) ◽

pp. 2291-2294

Keyword(s):

Image Processing ◽

Deep Learning ◽

Face Recognition ◽

Stereo Vision ◽

Large Scale ◽

Online Shopping ◽

Pressure Sensors ◽

Rfid Tags ◽

Weight Measurement ◽

The People

In recent years, shopping experiences are becoming more advanced. These include the attempts of market shelves as well as the currently booming online shopping. Online shopping has a better convenience but not yet accepted on a large scale by many people. Retail shops still retain greater response by the users and thus the retailers are moving towards an attempt of cashier-less shopping. A major problem of retail shops is that the people have crunch-time for shopping and cannot afford the waiting time at the checkout counters. Addressing this problem, we have developed a shopping style which saves time of checkout and also the time of maintaining social distancing queues. This research paper presents a stereo vision-based AI system which is useful to monitor the customers while shopping and also the items which are added or replaced in the virtual cart. The customers can directly walk out of the store after shopping and the final order cost of the shopping will be evaluated. This amount will be charged to the customer’s account. The system makes sure that there are no errors made during the evaluation and there are no charges for products which are not brought home. To achieve all this, the system uses image processing, object detection and face recognition algorithms that are widely practiced at present. The system also uses sensors like RFID tags and pressure sensors for weight measurement and detection of products on the shelves.

Download Full-text

Some comments about observations and image processing of comet 29P/Schwassmann-Wachmann 1

International Astronomical Union Colloquium ◽

10.1017/s0252921100031493 ◽

1999 ◽

Vol 173 ◽

pp. 243-248

Author(s):

D. Kubáček ◽

A. Galád ◽

A. Pravda

Keyword(s):

Image Processing ◽

Large Scale ◽

Harvard College ◽

Oak Ridge ◽

Large Scale Structures ◽

Short Period Comet ◽

Short Period ◽

Scale Structures ◽

Harvard College Observatory ◽

Period Comet

AbstractUnusual short-period comet 29P/Schwassmann-Wachmann 1 inspired many observers to explain its unpredictable outbursts. In this paper large scale structures and features from the inner part of the coma in time periods around outbursts are studied. CCD images were taken at Whipple Observatory, Mt. Hopkins, in 1989 and at Astronomical Observatory, Modra, from 1995 to 1998. Photographic plates of the comet were taken at Harvard College Observatory, Oak Ridge, from 1974 to 1982. The latter were digitized at first to apply the same techniques of image processing for optimizing the visibility of features in the coma during outbursts. Outbursts and coma structures show various shapes.

Download Full-text

Multi Disease-Prediction Framework Using Hybrid Deep Learning: An Optimal Prediction Model (Preprint)

10.2196/preprints.22865 ◽

2020 ◽

Author(s):

Anusha Ampavathi ◽

Vijaya Saradhi T

Keyword(s):

Feature Extraction ◽

Big Data ◽

Deep Learning ◽

Weight Function ◽

Optimization Algorithm ◽

Large Scale ◽

Heuristic Algorithms ◽

Disease Prediction ◽

Health Care Decisions ◽

Proposed Model

UNSTRUCTURED Big data and its approaches are generally helpful for healthcare and biomedical sectors for predicting the disease. For trivial symptoms, the difficulty is to meet the doctors at any time in the hospital. Thus, big data provides essential data regarding the diseases on the basis of the patient’s symptoms. For several medical organizations, disease prediction is important for making the best feasible health care decisions. Conversely, the conventional medical care model offers input as structured that requires more accurate and consistent prediction. This paper is planned to develop the multi-disease prediction using the improvised deep learning concept. Here, the different datasets pertain to “Diabetes, Hepatitis, lung cancer, liver tumor, heart disease, Parkinson’s disease, and Alzheimer’s disease”, from the benchmark UCI repository is gathered for conducting the experiment. The proposed model involves three phases (a) Data normalization (b) Weighted normalized feature extraction, and (c) prediction. Initially, the dataset is normalized in order to make the attribute's range at a certain level. Further, weighted feature extraction is performed, in which a weight function is multiplied with each attribute value for making large scale deviation. Here, the weight function is optimized using the combination of two meta-heuristic algorithms termed as Jaya Algorithm-based Multi-Verse Optimization algorithm (JA-MVO). The optimally extracted features are subjected to the hybrid deep learning algorithms like “Deep Belief Network (DBN) and Recurrent Neural Network (RNN)”. As a modification to hybrid deep learning architecture, the weight of both DBN and RNN is optimized using the same hybrid optimization algorithm. Further, the comparative evaluation of the proposed prediction over the existing models certifies its effectiveness through various performance measures.

Download Full-text