Three-dimensional Classification Research Articles

CNN-based 3D object classification using Hough space of LiDAR point clouds

With the wide application of Light Detection and Ranging (LiDAR) in the collection of high-precision environmental point cloud information, three-dimensional (3D) object classification from point clouds has become an important research topic. However, the characteristics of LiDAR point clouds, such as unstructured distribution, disordered arrangement, and large amounts of data, typically result in high computational complexity and make it very difficult to classify 3D objects. Thus, this paper proposes a Convolutional Neural Network (CNN)-based 3D object classification method using the Hough space of LiDAR point clouds to overcome these problems. First, object point clouds are transformed into Hough space using a Hough transform algorithm, and then the Hough space is rasterized into a series of uniformly sized grids. The accumulator count in each grid is then computed and input to a CNN model to classify 3D objects. In addition, a semi-automatic 3D object labeling tool is developed to build a LiDAR point clouds object labeling library for four types of objects (wall, bush, pedestrian, and tree). After initializing the CNN model, we apply a dataset from the above object labeling library to train the neural network model offline through a large number of iterations. Experimental results demonstrate that the proposed method achieves object classification accuracy of up to 93.3% on average.

Multi-channel Joint Sparse Learning Model for Non-rigid Three-dimensional Object Classification

Abstract In order to solve the issues of inadequate feature description and inefficient feature learning model existing in current classification methods, this article proposes a multi-channel joint sparse learning model for three-dimensional (3D) non-rigid object classification. First, the authors adopt a multi-level measurement of intrinsic properties to create complementary shape descriptors. Second, they build independent and informative bag of features (BoF) by embedding these shape descriptors into the visual vocabulary space. Third, a max-dependency and min-redundancy criterion is applied for optimal feature filtering on each BoF dictionary based on mutual information; meanwhile, each dictionary is learned and weighted according to its contribution to the classification task, and then a compact multi-channel joint sparse learning model is constructed. Finally, the authors train the joint sparse learning model followed by a Softmax classifier to implement efficient shape classification. The experimental results show that the proposed method has stronger feature representation ability and promotes greatly the discrimination of sparse coding coefficients. Thus, the promising classification performance and the powerful robustness can be obtained compared to the state-of-the-art methods.

Improved 3D tracking and automated classification of rodents' behavioral activity using depth-sensing cameras.

Analysis of rodents' behavior/activity is of fundamental importance in many research fields. However, many behavioral experiments still rely on manual scoring, with obvious problems in reproducibility. Despite important advances in video-analysis systems and computational ethology, automated behavior quantification is still a challenge. The need for large training datasets, background stability requirements, and reduction to two-dimensional analysis (impairing full posture characterization), limit their use. Here we present a novel integrated solution for behavioral analysis of individual rats, combining video segmentation, tracking of body parts, and automated classification of behaviors, using machine learning and computer vision methods. Low-cost depth cameras (RGB-D) are used to enable three-dimensional tracking and classification in dark conditions and absence of color contrast. Our solution automatically tracks five anatomical landmarks in dynamic environments and recognizes seven distinct behaviors, within the accuracy range of human annotations. The developed free software was validated in experiments where behavioral differences between Wistar Kyoto and Wistar rats were automatically quantified. The results reveal the capability for effective automated phenotyping. An extended annotated RGB-D dataset is also made publicly available. The proposed solution is an easy-to-use tool, with low-cost setup and powerful 3D segmentation methods (in static/dynamic environments). The ability to work in dark conditions means that natural animal behavior is not affected by recording lights. Furthermore, automated classification is possible with only ~30 minutes of annotated videos. By creating conditions for high-throughput analysis and reproducible quantitative measurements of animal behavior experiments, we believe this contribution can greatly improve behavioral analysis research.

Three-dimensional quantification and classification approach for angularity and surface texture based on surface triangulation of reconstructed aggregates

Angularity and surface texture of aggregates are important morphological characteristics, which have significant effects on the micromechanical responses of asphalt composites. A novel approach to quantify and classify the complexity of aggregate angularity and surface texture is proposed from a statistical perspective in this study. The methodology consists of three main steps, as follows: 1) the aggregate is three-dimensional (3D) reconstructed, and the model surface is triangulated into facets to conduct clustering for aggregate angles evaluation; 2) consequently, a reference plane is determined for each facet cluster to quantify the surface texture of a surface area which overlaps the cluster, and thus the surface texture is quantified; and 3) aggregates are classified by using the distribution characteristics of the angularity and surface texture respectively. Based on the quantification with the presented approach, 275 grains were classified by the angularity and surface texture. Results indicate the benefit of the proposed method in accurate quantification and classification of aggregate angularity and surface texture, which facilitate the bridging of the gap between aggregate morphology and micromechanical performance of asphalt composites.

Classification systems in adolescent idiopathic scoliosis revisited: Is a three-dimensional classification needed?

Classification systems for adolescent idiopathic scoliosis (AIS) have been in existence since the Schulthess system of 1905. Despite the numerous schema that have evolved over the last 115 years, little has changed from the original system based on the location of the coronal plane apex of the curves. Attempts at adding the sagittal plane, axial plane and shoulder balance to the system has generally yielded unscientific schemas or unwieldy numbers of variables within the scheme. The fundamental flaw with all these classifications is that they are based on two-dimensional imaging. The introduction of 3-D imaging like EOS and surface topography studies allow us an entirely novel perception of the spinal orientation in space. Thus the 3-D classifications that have emerged does not necessarily mean adding on Cartesian co-ordinates to the existing systems, but a far more comprehensive, yet simplistic view of the spinal deformity. Evidently, we are far from fully establishing all the variables and potentials of such schemas. Current modalities of 3-D imaging and evaluation are largely in the research domain and have not yet reached the clinical practice stage. The clinical utility of such 3-D classifications is also conjectural at present. But it is eminently possible that in the foreseeable future scoliosis classifications would cease to appear and be applied as they are today.

A Framework of User Classification Model of Online User Innovation Communities Based on User Innovation Value

The online user innovation communities (OUICs) provide a platform for users to participate in innovation activities, share product experience, and interact with each other. Different types of users have different innovation values. Classification of users can facilitate better community management and maximize the value of users. In the existing user classification research, the analysis of user innovation value is not in-depth enough, and the professional knowledge of each stage of product development possessed by users is not fully considered, which is the key to maximize the value of user innovation. Therefore, we comprehensively consider the user’s innovation capability, professional capability, influence capability and active capability, and propose a measurement method of user innovation value (4C). And based on user innovation value, a framework of a three-dimensional user classification model (TIS) was proposed combined with the product development stages and topics that users participated in, which contains topics, innovation value, and stage. It can enable companies to introduce high-innovation value users with different professional knowledge backgrounds at different stages of product development, and give full play to their innovation value, which is conducive to improving the product innovation performance of companies.

Frontal and sagittal imbalance in patients with adolescent idiopathic deformity.

The human spinal column underwent many significant changes over the 4.5 million years of our ancestral bipedalism. The main change, however, came with acquiring multiple curvatures in the sagittal plane. This alteration seems to have exposed a weakness in our body's keystone and made us susceptible to thus far unbeknown problems of the spine because it has been noted that idiopathic scoliosis has not been observed in other primates. Adolescent idiopathic scoliosis (AIS) is a three-dimensional deformity of the spine causing an imbalance of the trunk as it increases in magnitude. A scoliotic curve comprises three components, the coronal, sagittal, and axial so that each curve can affect the global balance of the body differently. Patients with significant scoliotic deformities often find themselves at a biomechanical disadvantage when it comes to energy expenditure and keeping an upright stance. The pioneers of scoliosis research recognized the need for describing and quantifying deformity to better understand it, so they first translated clinical measurements to radiographs and built from there. The development of concepts like defining a curve by its end vertebrae and measuring its magnitude, assessing global spinal balance, describing the stable zone, and pinpointing the stable vertebra all followed suit. The importance of sagittal balance and restoring sagittal parameters during treatment was emphasized. In a quest to bring order to chaos, some tried to classify various scoliotic curve types. These classifications helped steer treatment decisions but were found lacking in many aspects. So far, a widely accepted three-dimensional classification of scoliosis still does not exist. This review aims to provide the reader with an overview of the development of balance and imbalance concepts in scoliosis.

Dollarization in Central and Eastern European countries: essence, genesis, interrelation

The paper is devoted to scientific-theoretical and econometric analysis of dollarization on the example of Central and Eastern European countries. The views of scientists on the essence of the investigated category are generalized; the presence of high degree pluralism, particularly among Ukrainian researchers is determined. The authors’ definition, according to which the necessary conditions for dollarization in the economy such as: the recognition by residents the ability of foreign currency to perform one or more national money’s functions and systematic use of such currency in cash and/or non-cash circulation is given. Three-dimensional classification of dollarization, consistent with the modern money theory, according to the criteria of penalty, state recognition and performance of specific money’s functions is presented. Genesis and manifestations of inferred types are investigated. The global level of commitment to the U.S. dollar as the most common settlement and reserve currency is defined according to the global statistics of cross-border payments and the currency composition of the official central banks’ reserves. Methodological approaches for determining the level of dollarization of the economy, the most common of which – dollarization of money supply according to IMF and external debt–to–GDP ratio illustrated by statistics of 9 Central and Eastern European countries with own currencies are analyzed. It is determined that the high level of dollarization indicates the substitutionary competition between national and foreign money, where the latter are considered as a key factor for entering the global financial and commodity markets. According to the results of construction, estimation and re-estimation of BVAR dollarization model of Ukraine, Belarus, Poland and Russia, the significant causal relationship between the share of the currency component in the money supply of Russia and Poland and the corresponding indicators of Belarus and Ukraine is proved. This substantiates the previous conclusions about the unity of the dollarization evolutionary logic in historically close economies. The need for further empirical research of provoking and restraining factors of dollarization in Central and Eastern European countries is emphasized.

Three-Dimensional Evaluation and Classification of the Anatomy Variations of Vertebral Artery at the Craniovertebral Junction in 120 Patients of Basilar Invagination and Atlas Occipitalization.

Patients with basilar invagination and atlas occipitalization usually present abnormal anatomy of the vertebral arteries (VAs) at the craniovertebral junction (CVJ). To describe and further classify different types of VA variations at the CVJ with 3D visualization technology. One hundred twenty patients with basilar invagination and atlas occipitalization who had undergone 3-dimensional computed tomographic angiography (3D-CTA) were retrospectively studied. Imaging data were processed via the separating, fusing, opacifying, and false-coloring-volume rendering technique. Abnormal anatomy of the VA at the CVJ was categorized and related anatomic parameters were measured. Seven different types were classified. Type I, VAs enter the cranium after leaving VA groove on the posterior arch of atlas (26.7% of 240 sides); Type II, VAs enter an extraosseous canal created in the assimilated atlas lateral mass-occipital condyle complex before reaching the cranium (53.3%); Type III, VA courses above the axis facet or curves below the atlas lateral mass then enter the cranium (11.7%); Type IV, VAs enter the spinal canal under the axis lamina (1.3%); Type V, high-riding VA (31.3%); Type VI, fenestrated VA (2.9%); Type VII, absent VA (4.2%). Distance from the canal of Type II VA to the posterior facet surface of atlas lateral mass (5.51±2.17 mm) means a 3.5-mm screw can be safely inserted usually. Shorter distance from the midline (13.50±4.35) illustrates potential Type III VA injury during exposure. Decreased height and width of axis isthmus in Type V indicate increased VA injury risks. Seven types of VA variations were described, together with valuable information helpful to minimize VA injury risk intraoperatively.

Improving hospital pharmacy inventory management using data segmentation.

This case study describes the development and empirical validation of an easy-to-implement practical framework for improving hospital pharmacy inventory management. Research suggests various inventory optimization models, which can lead to cost reductions while maintaining adequate service levels; however, they are facing limited adoption in healthcare settings. The main barriers appear to be the high effort and complexity of implementation, the dependence on data that are not available or might not be in the right form, and the one-size-fits-all approach often followed without addressing healthcare sector-specific particularities. A research framework was developed by adapting relevant inventory models to the healthcare context using the concept of data segmentation on the basis of a three-dimensional classification of hospital pharmacy inventory items based on their relative importance, clinical criticality, and consumption pattern. Suitable replenishment policies were assigned to high-impact classes, and an integrated performance-measurement component assesses the framework's effectiveness. The suggested approach was implemented and empirically tested at the pharmacy of a large public hospital using longitudinal data. The results demonstrate substantial improvements with respect to all of the selected key performance indicators and translate into inventory cost savings due to reduced stockholding costs and better synchronization of inventories to demand. Use of standard software functionalities combined with targeted data segmentation efforts significantly improves hospital pharmacy inventory cost performance.

A User Interface for Optimizing Radiologist Engagement in Image Data Curation for Artificial Intelligence.

To delineate image data curation needs and describe a locally designed graphical user interface (GUI) to aid radiologists in image annotation for artificial intelligence (AI) applications in medical imaging. GUI components support image analysis toolboxes, picture archiving and communication system integration, third-party applications, processing of scripting languages, and integration of deep learning libraries. For clinical AI applications, GUI components included two-dimensional segmentation and classification; three-dimensional segmentation and quantification; and three-dimensional segmentation, quantification, and classification. To assess radiologist engagement and performance efficiency associated with GUI-related capabilities, image annotation rate (studies per day) and speed (minutes per case) were evaluated in two clinical scenarios of varying complexity: hip fracture detection and coronary atherosclerotic plaque demarcation and stenosis grading. For hip fracture, 1050 radiographs were annotated over 7 days (150 studies per day; median speed: 10 seconds per study [interquartile range, 3-21 seconds per study]). A total of 294 coronary CT angiographic studies with 1843 arteries and branches were annotated for atherosclerotic plaque over 23 days (15.2 studies [80.1 vessels] per day; median speed: 6.08 minutes per study [interquartile range, 2.8-10.6 minutes per study] and 73 seconds per vessel [interquartile range, 20.9-155 seconds per vessel]). GUI-component compatibility with common image analysis tools facilitates radiologist engagement in image data curation, including image annotation, supporting AI application development and evolution for medical imaging. When complemented by other GUI elements, a continuous integrated workflow supporting formation of an agile deep neural network life cycle results.Supplemental material is available for this article.© RSNA, 2019.

Joint 3D localization and classification of space debris using a multispectral rotating point spread function.

We consider the problem of joint three-dimensional (3D) localization and material classification of unresolved space debris using a multispectral rotating point spread function (RPSF). The use of RPSF allows one to estimate the 3D locations of point sources from their rotated images acquired by a single 2D sensor array, since the amount of rotation of each source image about its x, y location depends on its axial distance z. Using multispectral images, with one RPSF per spectral band, we are able not only to localize the 3D positions of the space debris but also classify their material composition. We propose a three-stage method for achieving joint localization and classification. In stage 1, we adopt an optimization scheme for localization in which the spectral signature of each material is assumed to be uniform, which significantly improves efficiency and yields better localization results than possible with a single spectral band. In stage 2, we estimate the spectral signature and refine the localization result via an alternating approach. We process classification in the final stage. Both Poisson noise and Gaussian noise models are considered, and the implementation of each is discussed. Numerical tests using multispectral data from NASA show the efficiency of our three-stage approach and illustrate the improvement of point source localization and spectral classification from using multiple bands over a single band.

Applications of Game Theory in Project Management: A Structured Review and Analysis

This paper provides a structured literature review and analysis of using game theory to model project management scenarios. We select and review thirty-two papers from Scopus, present a complex three-dimensional classification of the selected papers, and analyse the resultant citation network. According to the industry-based classification, the surveyed literature can be classified in terms of construction industry, ICT industry or unspecified industry. Based on the types of players, the literature can be classified into papers that use government-contractor games, contractor–contractor games, contractor-subcontractor games, subcontractor–subcontractor games or games involving other types of players. Based on the type of games used, papers using normal-form non-cooperative games, normal-form cooperative games, extensive-form non-cooperative games or extensive-form cooperative games are present. Also, we show that each of the above classifications plays a role in influencing which papers are likely to cite a particular paper, though the strongest influence is exerted by the type-of-game classification. Overall, the citation network in this field is sparse, implying that the awareness of authors in this field about studies by other academics is suboptimal. Our review suggests that game theory is a very useful tool for modelling project management scenarios, and that more work needs to be done focusing on project management in ICT domain, as well as by using extensive-form cooperative games where relevant.

New guidelines for graftless rehabilitation of the posterior maxilla

Background The maxillary sinus anatomy varies according to age, population groups, sex and absence of teeth. Three-dimensional reconstruction and simulation of the maxillary sinus allows its classification into different categories based on the morphology of the inferior wall. Rehabilitation protocols for the posterior maxilla range from short implants, sinus graft, to zygomatic implants. This research proposes the creation of guidelines following three-dimensional classification of the maxillary sinus. Aim/Hypothesis It is possible to perform graftless rehabilitation of the posterior edentulous maxilla in a predictable way using dental implants. Material and Methods We performed a literature review on the treatment alternatives concerning the maxillary sinus anatomical categories. Following different sinus classifications, the authors proposed a decision matrix to categorize the different anatomical variations of the maxillary sinus. We applied specific rehabilitation models considering the different types of implant length, diameter and angulation versus the different three-dimensional categories and used real clinical cases to demonstrate the viability of the guideline for each category. Results Concerning the different sinus anatomy, there are six different classifications. The authors provided different graftless rehabilitation alternatives considering the use of standard implants (straight and tilted) and using the maxilla, together with the pterygoid or zygomatic bones for anchorage. The graftless rehabilitation alternatives are illustrated by clinical cases and discussed in the light of contemporary scientific literature and data from a series of cases. Considering the series of cases, the short term outcome was favorable with a high survival rate, low number of complications and patients’ reported satisfaction with function and aesthetics registered in all types of rehabilitations irrespective of the bone used for implant anchorage (maxilla only or in combination with pterygoid and zygomatic bone). Conclusion and Clinical Implications The commercial availability of grafting materials forced its indications spreading across a wider array of clinical presentations where otherwise conventional dental implants may be enough. It is possible to perform graftless rehabilitation of the edentulous posterior maxilla in cases where today many clinicians use bone substitutes as an alternative. The proposed clinical guidelines provide a treatment alternative using the maxilla, pterygoid or zygomatic bones for implant anchorage.

Morphology-dependent electromagnetic wave absorbing properties of iron-based absorbers: one-dimensional, two-dimensional, and three-dimensional classification

Owing to the fast development of wireless techniques at the high-frequency range, the electromagnetic interference problem has been of increasing significance and attracting global attention. It is urgent to develop efficient microwave absorbing materials to attenuate the harmful electromagnetic wave. Iron and Fe-based composites are advantageous in the low-cost and attractive magnetic properties, so they have been widely studied in microwave absorption. This review focuses on the latest advances in nanostructured Fe-based materials including nanostructured iron, Fe/C (carbon nanotubes, nanofibers, nanocapsules, etc.), Fe/semiconductor (TiO2, MnO2, ZnO, SiO2, MoS2, etc.), Fe/polymer (polyaniline and polypyrrole), FeCo alloy, etc. However, most of these Fe-based materials suffer from the poor impedance matching and oxidation, which seriously impede their implementation as high-performance microwave absorbing materials. In this review, the main synthesis and modification methods, as well as the practical performance of Fe-based microwave absorbing materials are discussed. Moreover, challenges and perspectives of Fe-based composites for further development in microwave absorbing materials are proposed.

A Three-dimensional Classification System of Second Language Self-repair

Self-repair is the change(s) a speaker makes to his/her ongoing speech due to any concern in talk. Previous psycholinguistic taxonomies of self-repair regard it as manifestations of individual speech production problems. This study questions this stand by investigating the self-repair behaviour of English language learners from two secondary schools in China, examining their task discourse and stimulated-recall comments. The results enables a three-dimensional classification system that analyses self-repair with reference to 1) the change(s) involved in a self-repair that is reflected in a certain domain of language, 2) the strategy employed to make the change(s) , and 3) the problem that prompts a speaker to self-repair. Analysis of the problems underlying the occurrences of self-repair demonstrates that L2 learners use self-repair to deal with problems in the interactions and the communicative contexts, in addition to their production problems.

Application of three-dimensional CT and image classification in percutaneous vertebraplasty for osteoporotic vertebral compression fractures

To explore the application of three-dimensional CT and image classification in the treatment of osteoporotic vertebral compression fracture(OVCFs) by percutaneous vertebroplasty(PVP). A total of 90 patients with OVCFs who were treated with PVP in Linqu People's Hospital of Shandong Province from April 2016 to March 2018 were selected as subjects. There were 31 males and 59 females, aged from 63 to 84 years old. Bone mineral density measurements were performed in all patients to confirm the presence of osteoporosis and imaging examinations were performed to confirm the presence of vertebral fractures. The fracture area was determined by MRI fat surpressed image before operation and three-dimensional modeling was performed to calculate the volume of fracture area. Three dimensional CT imaging of bone cement in fracture area was performed after PVP and the volume ratio of bone cement in fracture area was calculated by computer aided design software, by which patients were divided into groups for study. Forty-one patients whose volume ratio of bone cement in fracture area less than 50% are control group and the rest of 90 patients are observation group. Visual analogue scale (VAS) and Oswestry Disability Index(ODI) were collected in two groups before operation and 1 day, 3 months after operation. The amount of bone cement was recorded after operation. All operations were successful. There were 3 cases of cement leakage in control group and 4 cases in observation group. All patients had no obvious clinical symptoms. After continuous observation and follow-up for 3 months, no complications such as adjacent vertebral fracture, infection, bone cement displacement were found. There was no significant difference in bone cement doses and bone cement leakage between two groups(P>0.05). There was no significant difference in preoperative VAS and ODI between two groups(P>0.05). All VAS and ODI obviously decreased(P<0.05) at 1 day after operation and in observation group the decrease was more significant (P<0.05). At 3 months after operation there was no significant difference between two groups. This may have been due to basically healing of vertebral fractures at 3 months after surgery and the pain was no longer significantly related. PVP can significantly improve clinical symptoms of OVCFs and bone cement filling in fracture area is the key to the short-term effect of PVP.

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.

A galaxy classification grid that better recognises early-type galaxy morphology

A modified galaxy classification scheme for local galaxies is presented. It builds upon the Aitken-Jeans nebula sequence, by expanding the Jeans-Hubble tuning fork diagram, which itself contained key ingredients from Curtis and Reynolds. The two-dimensional grid of galaxy morphological types presented here, with elements from de Vaucouleurs' three-dimensional classification volume, has an increased emphasis on the often overlooked bars and continua of disc sizes in early-type galaxies - features not fully captured by past tuning forks, tridents, or combs. The grid encompasses nuclear discs in elliptical (E) galaxies, intermediate-scale discs in ellicular (ES) galaxies, and large-scale discs in lenticular (S0) galaxies, while the E4-E7 class is made redundant given that these galaxies are lenticular galaxies. Today, these disc structures continue to be neglected, or surprise researchers, perhaps partly due to our indoctrination to galaxy morphology through the tuning fork diagram. To better understand the past and present classification schemes - whose origins reside in solar/planetary formation models - a holistic overview is given. This provides due credit to some of the lesser known pioneers, presents some rationale for the grid, and reveals the incremental nature of, and some of the lesser known connections in, the field of galaxy morphology.

Three-Dimensional Classification Structure-Activity Relationship Analysis Using Convolutional Neural Network.

Quantitative structure-activity relationship (QSAR) techniques, especially those that possess three-dimensional attributes, such as the comparative molecular field analysis (CoMFA), are frequently used in modern-day drug design and other related research domains. However, the requirement for accurate alignment of compounds in CoMFA increases the difficulties encountered in its use. This has led to the development of several techniques-such as VolSurf, Grid-independent descriptors (GRIND), and Anchor-GRIND-which do not require such an alignment. We propose a technique to construct the prediction model that uses molecular interaction field grid potentials as inputs to convolutional neural network. The proposed model has been found to demonstrate higher accuracy compared to the conventional descriptor-based QSAR models as well as Anchor-GRIND techniques. In addition, the method is target independent, and is capable of providing useful information regarding the importance of individual atoms constituting the compounds contained in the chemical dataset used in the proposed analysis. In view of these advantages, the proposed technique is expected to find wide applications in future drug-design operations.