Accurate Object Classification Research Articles

Realtime Single-Shot Refinement Neural Network With Adaptive Receptive Field for 3D Object Detection From LiDAR Point Cloud

Object detection plays an important role in autonomous driving systems. LiDAR is widely used in autonomous driving vehicles and robots as a sensor for environmental perception. Recently, the development of computational power and deep learning technology makes it possible to classify and locate objects from LiDAR point cloud in a single end-to-end learnable network. However, objects are sparsely distributed in large point cloud field, and are always been partly scanned by LiDAR, which pose a challenge for accurate and rapid object positioning and classification from the raw point cloud. In this paper, we introduce a new single-shot refinement neural network for fast and accurate 3D object detection from the raw LiDAR point cloud. Firstly, we exploit self-attention mechanism in main object detection branch to enhance object feature representation. Secondly, we apply deformable convolution for learning adaptive receptive fields to fully capture the features of rotating and partially visible objects. Thirdly, an object refinement branch is introduced to produce a finer regression of objects upon the primary estimation from the main detection branch. All proposed modules have been proven to effectively improve the accuracy of object detection. Our method is evaluated on KITTI 3D detection benchmark and achieves state-of-the-art results while maintains real-time efficiency. Furthermore, real-time test in autonomous driving vehicle demonstrates that our method is robust to 16 channels LiDAR and can meet the demands of accuracy, efficiency, and visibility of object detection in various urban scenarios.

A New Method Based on Deep Convolutional Neural Networks for Object Detection and Classification

Accurate object detection and classification has a broad application in industrial tasks, such as fabric defect and invoice detection. Previous state-of-the-art methods such as SSD and Faster-RCNN usually need to carefully adjust anchor box related hyper parameters and have poor performance in special fields with large object size/ratio variations and complex background texture. In this study, we proposed a new accurate, robust, and anchor-free method to handle automatic object detection and classification problems. First, we used the feature pyramid network (FPN), to merge the feature maps of different scales of features extracted from a convolutional neural network (CNN), which allowed easy and robust multi-scale feature fusion. Second, we built two subnets to generate candidate region proposals from the FPN outputs. followed by another CNN that determined the categories of the proposed regions from the two subnets.

Development of accurate classification of heavenly bodies using novel machine learning techniques

The heavenly bodies are objects that swim in the outer space. The classification of these objects is a challenging task for astronomers. This article presents a novel methodology that enables an efficient and accurate classification of cosmic objects (3 classes) based on evolutionary optimization of classifiers. This research collected the data from Sloan Digital Sky Survey database. In this work, we are proposing to develop a novel machine learning model to classify stellar spectra of stars, quasars and galaxies. First, the input data are normalized and then subjected to principal component analysis to reduce the dimensionality. Then, the genetic algorithm is implemented on the data which helps to find the optimal parameters for the classifiers. We have used 21 classifiers to develop an accurate and robust classification with fivefold cross-validation strategy. Our developed model has achieved an improvement in the accuracy using nineteen out of twenty-one models. We have obtained the highest classification accuracy of 99.16%, precision of 98.78%, recall of 98.08% and F1-score of 98.32% using evolutionary system based on voting classifier. The developed machine learning prototype can help the astronomers to make accurate classification of heavenly bodies in the sky. Proposed evolutionary system can be used in other areas where accurate classification of many classes is required.

Clustering Nuclear Magnetic Resonance: Machine learning assistive rapid two‐dimensional relaxometry mapping

AbstractLow‐field nuclear magnetic resonance (NMR) relaxometry is an attractive approach from point‐of‐care testing medical diagnosis to in situ oil‐gas exploration. One of the problems, however, is the inherently long relaxation time of the (liquid) samples, (and hence low signal‐to‐noise ratio) which causes unnecessarily long repetition time. In this work, a new class of methodology is presented for rapid and accurate object classification using NMR relaxometry with the aid of machine learning. It is demonstrated that the sensitivity and specificity of the classification is substantially improved with higher order of (pseudo)‐dimensionality (e.g., 2D‐ or multi‐dimensional). This new methodology (termed as ‘Clustering NMR’) may be extremely useful for rapid and accurate object classification (in less than a minute) using low‐field NMR.

Hyperspectral Image Classification Based on Stacked Contractive Autoencoder Combined With Adaptive Spectral-Spatial Information

Hyperspectral image (HSI) contain abundant spectral and spatial information, enabling the accurate classification of ground objects. However, many existing machine learning methods have poor performance, and some existing CNN-based methods require high computational power, which considerably limits their real-world applications. To address these issues, in this paper, we propose an alternative HSI classification method based on the stacked contractive autoencoder (SCAE) and adaptive spectral-spatial information to improve the accuracy of HSI classification. Specifically, the non-subsampled shearlet transform (NSST) with the guided filtering (NG) enhances spatial structure information. Subsequently, we present an adaptive spatial information extraction method to extract the spatial information of pixels. Furthermore, we propose an HSI classification network, called SCAE-LR, for feature extraction and classification. The SCAE is implemented to extract the adaptive spectral-spatial feature, and a logistic regression (LR) layer is employed for classification. Extensive experiments on the Indian Pines data set and the Pavia University data set demonstrate the superior performance of our method.

Anti-noise diffractive neural network for constructing an intelligent imaging detector array.

To develop an intelligent imaging detector array, a diffractive neural network with strong robustness based on the Weight-Noise-Injection training is proposed. According to layered diffractive transformation under existing several errors, an accurate and fast object classification can be achieved. The fact that the mapping between the input image and the label in Weight-Noise-Injection training mode can be learned, means that the prediction of the optical network being insensitive to disturbances so as to improve its noise resistance remarkably. By comparing the accuracy under different noise conditions, it is verified that the proposed model can exhibit a higher accuracy.

Exo-atmospheric infrared objects classification based on dual-channel LSTM network

In order to realize accurate classification of Exo-atmospheric objects, a new framework which depends on dual-channel LSTM (Long Short-Term Memory) is presented in this paper. The influencing factors of Exo-atmospheric objects infrared imaging are analyzed and the simulation model of infrared image sequence is studied, which establishes the data foundation for the following research. After analyzing the characteristics of exo-atmospheric objects’ gray series, the input sequences are down-sampled and truncated separately at first, and then fed into corresponding LSTM channels to extract the global and local features respectively. Finally, the features of two channels are fused to classify the targets. Experimental results show that the proposed framework achieves better classification performance, and due to its good computing structure, it also has good practicability.

Deep Learning-Based Object Classification and Position Estimation Pipeline for Potential Use in Robotized Pick-and-Place Operations

Accurate object classification and position estimation is a crucial part of executing autonomous pick-and-place operations by a robot and can be realized using RGB-D sensors becoming increasingly available for use in industrial applications. In this paper, we present a novel unified framework for object detection and classification using a combination of point cloud processing and deep learning techniques. The proposed model uses two streams that recognize objects on RGB and depth data separately and combines the two in later stages to classify objects. Experimental evaluation of the proposed model including classification accuracy compared with previous works demonstrates its effectiveness and efficiency, making the model suitable for real-time applications. In particular, the experiments performed on the Washington RGB-D object dataset show that the proposed framework has 97.5% and 95% fewer parameters compared to the previous state-of-the-art multimodel neural networks Fus-CNN, CNN Features and VGG3D, respectively, with the cost of approximately 5% drop in classification accuracy. Moreover, the inference of the proposed framework takes 66.11%, 32.65%, and 28.77% less time on GPU and 86.91%, 51.12%, and 50.15% less time on CPU in comparison to VGG3D, Fus-CNN, and CNN Features. The potential applicability of the developed object classification and position estimation framework was then demonstrated on an experimental robot-manipulation setup realizing a simplified object pick-and-place scenario. In approximately 95% of test trials, the system was able to accurately position the robot over the detected objects of interest in an automatic mode, ensuring stable cyclic execution with no time delays.

Fusion of Hyperspectral CASI and Airborne LiDAR Data for Ground Object Classification through Residual Network.

Modern satellite and aerial imagery outcomes exhibit increasingly complex types of ground objects with continuous developments and changes in land resources. Single remote-sensing modality is not sufficient for the accurate and satisfactory extraction and classification of ground objects. Hyperspectral imaging has been widely used in the classification of ground objects because of its high resolution, multiple bands, and abundant spatial and spectral information. Moreover, the airborne light detection and ranging (LiDAR) point-cloud data contains unique high-precision three-dimensional (3D) spatial information, which can enrich ground object classifiers with height features that hyperspectral images do not have. Therefore, the fusion of hyperspectral image data with airborne LiDAR point-cloud data is an effective approach for ground object classification. In this paper, the effectiveness of such a fusion scheme is investigated and confirmed on an observation area in the middle parts of the Heihe River in China. By combining the characteristics of hyperspectral compact airborne spectrographic imager (CASI) data and airborne LiDAR data, we extracted a variety of features for data fusion and ground object classification. Firstly, we used the minimum noise fraction transform to reduce the dimensionality of hyperspectral CASI images. Then, spatio-spectral and textural features of these images were extracted based on the normalized vegetation index and the gray-level co-occurrence matrices. Further, canopy height features were extracted from airborne LiDAR data. Finally, a hierarchical fusion scheme was applied to the hyperspectral CASI and airborne LiDAR features, and the fused features were used to train a residual network for high-accuracy ground object classification. The experimental results showed that the overall classification accuracy was based on the proposed hierarchical-fusion multiscale dilated residual network (M-DRN), which reached an accuracy of 97.89%. This result was found to be 10.13% and 5.68% higher than those of the convolutional neural network (CNN) and the dilated residual network (DRN), respectively. Spatio-spectral and textural features of hyperspectral CASI images can complement the canopy height features of airborne LiDAR data. These complementary features can provide richer and more accurate information than individual features for ground object classification and can thus outperform features based on a single remote-sensing modality.

A Sustainable Deep Learning Framework for Object Recognition Using Multi-Layers Deep Features Fusion and Selection

With an overwhelming increase in the demand of autonomous systems, especially in the applications related to intelligent robotics and visual surveillance, come stringent accuracy requirements for complex object recognition. A system that maintains its performance against a change in the object’s nature is said to be sustainable and it has become a major area of research for the computer vision research community in the past few years. In this work, we present a sustainable deep learning architecture, which utilizes multi-layer deep features fusion and selection, for accurate object classification. The proposed approach comprises three steps: (1) By utilizing two deep learning architectures, Very Deep Convolutional Networks for Large-Scale Image Recognition and Inception V3, it extracts features based on transfer learning, (2) Fusion of all the extracted feature vectors is performed by means of a parallel maximum covariance approach, and (3) The best features are selected using Multi Logistic Regression controlled Entropy-Variances method. For verification of the robust selected features, the Ensemble Learning method named Subspace Discriminant Analysis is utilized as a fitness function. The experimental process is conducted using four publicly available datasets, including Caltech-101, Birds database, Butterflies database and CIFAR-100, and a ten-fold validation process which yields the best accuracies of 95.5%, 100%, 98%, and 68.80% for the datasets respectively. Based on the detailed statistical analysis and comparison with the existing methods, the proposed selection method gives significantly more accuracy. Moreover, the computational time of the proposed selection method is better for real-time implementation.

Image-free classification of fast-moving objects using "learned" structured illumination and single-pixel detection.

Object classification generally relies on image acquisition and subsequent analysis. Real-time classification of fast-moving objects is a challenging task. Here we propose an approach for real-time classification of fast-moving objects without image acquisition. The key to the approach is to use structured illumination and single-pixel detection to acquire the object features directly. A convolutional neural network (CNN) is trained to learn the object features. The "learned" object features are then used as structured patterns for structured illumination. Object classification can be achieved by picking up the resulting light signals by a single-pixel detector and feeding the single-pixel measurements to the trained CNN. In our experiments, we show that accurate and real-time classification of fast-moving objects can be achieved. Potential applications of the proposed approach include rapid classification of flowing cells, assembly-line inspection, and aircraft classification in defense applications. Benefiting from the use of a single-pixel detector, the approach might be applicable for hidden moving object classification.

Analysis of HMAX Algorithm on Black Bar Image Dataset

An accurate detection and classification of scenes and objects is essential for interacting with the world, both for living beings and for artificial systems. To reproduce this ability, which is so effective in the animal world, numerous computational models have been proposed, frequently based on bioinspired, computational structures. Among these, Hierarchical Max-pooling (HMAX) is probably one of the most important models. HMAX is a recognition model, mimicking the structures and functions of the primate visual cortex. HMAX has already proven its effectiveness and versatility. Nevertheless, its computational structure presents some criticalities, whose impact on the results has never been systematically assessed. Traditional assessments based on photographs force to choose a specific context; the complexity of images makes it difficult to analyze the computational structure. Here we present a new, general and unspecific assessment of HMAX, introducing the Black Bar Image Dataset, a customizable set of images created to be a universal and flexible model of any ‘real’ image. Results: surprisingly, HMAX demonstrates a notable sensitivity also with a low contrast of luminance. Images containing a wider information pattern enhance the performances. The presence of textures improves performance, but only if the parameterization of the Gabor filter allows its correct encoding. In addition, in complex conditions, HMAX demonstrates good effectiveness in classification. Moreover, the present assessment demonstrates the benefits offered by the Black Bar Image Dataset, its modularity and scalability, for the functional investigations of any computational models.

AI on a chip.

Artificial intelligence (AI) has dramatically changed the landscape of science, industry, defence, and medicine in the last several years. Supported by considerably enhanced computational power and cloud storage, the field of AI has shifted from mostly theoretical studies in the discipline of computer science to diverse real-life applications such as drug design, material discovery, speech recognition, self-driving cars, advertising, finance, medical imaging, and astronomical observation, where AI-produced outcomes have been proven to be comparable or even superior to the performance of human experts. In these applications, what is essentially important for the development of AI is the data needed for machine learning. Despite its prominent importance, the very first process of the AI development, namely data collection and data preparation, is typically the most laborious task and is often a limiting factor of constructing functional AI algorithms. Lab-on-a-chip technology, in particular microfluidics, is a powerful platform for both the construction and implementation of AI in a large-scale, cost-effective, high-throughput, automated, and multiplexed manner, thereby overcoming the above bottleneck. On this platform, high-throughput imaging is a critical tool as it can generate high-content information (e.g., size, shape, structure, composition, interaction) of objects on a large scale. High-throughput imaging can also be paired with sorting and DNA/RNA sequencing to conduct a massive survey of phenotype-genotype relations whose data is too complex to analyze with traditional computational tools, but is analyzable with the power of AI. In addition to its function as a data provider, lab-on-a-chip technology can also be employed to implement the developed AI for accurate identification, characterization, classification, and prediction of objects in mixed, heterogeneous, or unknown samples. In this review article, motivated by the excellent synergy between AI and lab-on-a-chip technology, we outline fundamental elements, recent advances, future challenges, and emerging opportunities of AI with lab-on-a-chip technology or "AI on a chip" for short.

Single-shot spatially-localized NQR using field-dependent relaxation rates

Nuclear quadrupole resonance (NQR) is commonly used to characterize solid materials containing quadrupolar nuclei. For example, NQR is a promising technique for detecting plastic explosives and other forbidden substances as well as for authenticating pharmaceutical products. Spatially-resolved NQR measurements are of particular interest for enabling automated sample positioning, evaluation of sample heterogeneity, and chemometric authentication of objects. This paper proposes a rapid “single-shot” method for spatially-resolved NQR with the potential to benefit such applications. The proposed method takes advantage of the fact that certain NQR relaxation rates are field-dependent: the observed field dependence is used to convert relaxation time distributions measured in a static field gradient (estimated via Laplace inversion of time-domain data) into spatial distributions. The method was validated using 35Cl and 37Cl NQR of sodium chlorate and other compounds. Effective spatial resolution was also improved by using machine learning (ML) to classify the measured spatial distributions. In particular, experimental results demonstrate accurate ML-based classification of 3D-printed objects containing arbitrary binary distributions of sodium chlorate. Such distributions can thus be used as NQR-based “embedded barcodes” for authenticating high-value objects.

True-Color Three-Dimensional Imaging and Target Classification BASED on Hyperspectral LiDAR

True-color three-dimensional (3D) imaging exploits spatial and spectral information and can enable accurate feature extraction and object classification. The existing methods, however, are limited by data collection mechanisms when realizing true-color 3D imaging. We overcome this problem and present a novel true-color 3D imaging method based on a 32-channel hyperspectral LiDAR (HSL) covering a 431–751 nm spectral range. We conducted two experiments, one with nine-color card papers and the other with seven different colored objects. We used the former to investigate the effect of true-color 3D imaging and determine the optimal spectral bands for compositing true-color, and the latter to explore the classification potential based on the true-color feature using polynomial support vector machine (SVM) and Gaussian naive Bayes (NB) classifiers. Since using all bands of HSL will cause color distortions, the optimal spectral band combination for better compositing the true-color were selected by principal component analysis (PCA) and spectral correlation measure (SCM); PCA emphasizes the amount of information in band combinations, while SCM focuses on correlation between bands. The results show that the true-color 3D imaging can be realized based on HSL measurements, and three spectral bands of 466, 546, and 626 nm were determined. Comparing reflectance of the three selected bands, the overall classification accuracy of seven different colored objects was improved by 14.6% and 8.25% based on SVM and NB, respectively, classifiers after converting spectral intensities into true-color information. Overall, this study demonstrated the potential of HSL system in retrieving true-color and facilitating target recognition, and can serve as a guide in developing future three-channel or multi-channel true-color LiDAR.

Visual and Auditory Data Fusion for Energy-Efficient and Improved Object Recognition in Wireless Multimedia Sensor Networks

Automatic threat classification without human intervention is a popular research topic in wireless multimedia sensor networks (WMSNs) especially within the context of surveillance applications. This paper explores the effect of fusing audio-visual multimedia and scalar data collected by the sensor nodes in a WMSN for the purpose of energy-efficient and accurate object detection and classification. In order to do that, we implemented a wireless multimedia sensor node with video and audio capturing and processing capabilities in addition to traditional/ordinary scalar sensors. The multimedia sensors are kept in sleep mode in order to save energy until they are activated by the scalar sensors which are always active. The object recognition results obtained from video and audio applications are fused to increase the object recognition performance of the sensor node. Final results are forwarded to the sink in text format, and this greatly reduces the size of data transmitted in network. Performance test results of the implemented prototype system show that the fusing audio data with visual data improves automatic object recognition capability of a sensor node significantly. Since auditory data requires less processing power compared to visual data, the overhead of processing the auditory data is not high, and it helps to extend network lifetime of WMSNs.

Visual Object Recognition and Pose Estimation Based on a Deep Semantic Segmentation Network

In recent years, deep learning-based object recognition algorithms become emerging in robotic vision applications. This paper addresses the design of a novel deep learning-based visual object recognition and pose estimation system for a robot manipulator to handle random object picking tasks. The proposed visual control system consists of a visual perception module, an object pose estimation module, a data argumentation module, and a robot manipulator controller. The visual perception module combines deep convolution neural networks (CNNs) and a fully connected conditional random field layer to realize an image semantic segmentation function, which can provide stable and accurate object classification results in cluttered environments. The object pose estimation module implements a model-based pose estimation method to estimate the 3D pose of the target for picking control. In addition, the proposed data argumentation module automatically generates training data for training the deep CNN. Experimental results show that the proposed scene segmentation method used in the data argumentation module reaches a high accuracy rate of 97.10% on average, which is higher than other state-of-the-art segment methods. Moreover, with the proposed data argumentation module, the visual perception module reaches an accuracy rate over than 80% and 72% in the case of detecting and recognizing one object and three objects, respectively. In addition, the proposed model-based pose estimation method provides accurate 3D pose estimation results. The average translation and rotation errors in the three axes are all smaller than 0.52 cm and 3.95 degrees, respectively. These advantages make the proposed visual control system suitable for applications of random object picking and manipulation.

Enhanced Fusion of Deep Neural Networks for Classification of Benchmark High-Resolution Image Data Sets

Accurate land cover classification and detection of objects in high-resolution electro-optical remote sensing imagery (RSI) have long been a challenging task. Recently, important new benchmark data sets have been released which are suitable for land cover classification and object detection research. Here, we present state-of-the-art results for four benchmark data sets using a variety of deep convolutional neural networks (DCNN) and multiple network fusion techniques. We achieve 99.70%, 99.66%, 97.74%, and 97.30% classification accuracies on the PatternNet, RSI-CB256, aerial image, and RESISC-45 data sets, respectively, using the Choquet integral with a novel data-driven optimization method presented in this letter. The relative reduction in classification errors achieved by this data driven optimization is 25%–45% compared with the single best DCNN results.

Accurate object classification and detection by faster-RCNN

Recent advances in object detection algorithms include fast and faster RCNN which made the detection times comparatively low with high accuracy. In this work we verify the integrity of a proposed algorithm which uses RPN (Region proposal networks) and Fast RCNN (Region based Convolutional Neural Networks) for the detection. The RPN provides region proposals from that we give the ROI (Regions of Interest) as input to the RCNN network, it can be further merged into a single network by sharing their convolutional features to detect a specific object in a given image. As we use a unified network there is no need to get the ROI from an external network which makes this process cost free. We trained VGGnet with two different data sets PASCAL VOC 2012 and MS COCO on a low cost GPU and verified the accuracies while comparing the outputs with increasing number of region proposals. As we increased the number of proposals we observed a significant increase in the mAP (Mean Average Precision) value till 2000 proposals from where it reached saturation. Our results are compared with the state of the art algorithm with an increase of 1.2% in terms of mAP for 1800 proposals.

A Novel Bayesian Spatial–Temporal Random Field Model Applied to Cloud Detection From Remotely Sensed Imagery

With the fast advancement of remote sensing platforms and sensors, remotely sensed imagery (RSI) is increasingly being characterized by both high spatial resolution and high temporal resolution. How to efficiently use the rich spatial and temporal information in RSI for highly accurate object detection and classification is an important research question. Nevertheless, there is still a lack of a probabilistic framework that is capable of fully accounting for the spatial–temporal information in RSI for improved applications. In this paper, we present a Bayesian spatial–temporal random field model that constitutes a complete probabilistic framework for fully explaining the spatial–temporal correlation in RSI, leading to an enhanced object detection approach that is used for cloud detection from RSI. Under the Bayesian theorem, the posterior distribution of a label field is decomposed into the label prior, the data likelihood, the temporal label likelihood, and the temporal data likelihood. To address the difficulties in modeling the complex spatial–temporal correlation effect in the temporal data likelihood, a stochastic sampling approach is presented. Based on the maximum a posteriori approach, the posterior distribution is seamlessly integrated into the graph-cut optimization framework, and, therefore, the model optimization can be efficiently solved. The proposed algorithm is tested for cloud detection on both simulated and real RSIs and the results demonstrate that the proposed algorithm can effectively exploit the spatial–temporal information for achieving higher detection accuracy.