Pose Orientation Research Articles

Exploring the Anticancer Potential of Furanpydone A: A Computational Study on its Inhibition of MTHFD2 Across Diverse Cancer Cell Lines.

Cancer poses a significant global health challenge due to its high mortality rate and complex treatment strategies. Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2), which is notably overexpressed in various malignancies, represents a promising target for anticancer drug development. Furanpydone A, a new 4-hydroxy-2-pyridone alkaloid isolated from the endophytic fungus Arthrinium sp. GZWMJZ-606, has shown potent inhibitory activity against several cancer cell lines. This study provides the first computational evaluation of furanpydone A, focusing on its potential inhibition of MTHFD2 through molecular docking and 200 ns molecular dynamics (MD) simulations. Molecular docking revealed a binding free energy of -8.08 kcal/mol for furanpydone A, comparable to the control compound DS44960156 (-8.13 kcal/mol), indicating stable interactions with the MTHFD2 active site. MD simulations confirmed the structural stability of the furanpydone A-MTHFD2 complex, with RMSD values ranging from 1.5 to 2.9 Å, RMSF values below 4 Å, and a radius of gyration (Rg) of 26.7 Å. Furanpydone A maintained approximately four consistent hydrogen bonds throughout the simulation. Analysis of furanpydone A binding pose orientations and interactions with the MTHFD2 enzyme at 0 ns, 40 ns, 80 ns, 120 ns, 160 ns, and 200 ns revealed consistent and stable binding. MM-PBSA analysis showed a binding free energy (ΔGbind) of -23.57 ± 0.13 kcal/mol, with electrostatic and van der Waals interactions contributing significantly, suggesting competitive binding affinity to the control compound (-25.32 ± 0.11 kcal/mol). The contribution of individual amino acid residues, including key residues such as ARG43, TYR84, ASN87, LYS88, GLN132, and PRO314, indicated strong interactions that support the stability of the furanpydone A-MTHFD2 complex. ADMET predictions indicated that furanpydone A met key drug-likeness criteria and demonstrated good oral bioavailability, suitable distribution profile, minimal risk of drug-drug interactions, efficient elimination, and low toxicity potential. These findings suggest that furanpydone A is a promising candidate for cancer treatment, warranting further in vitro and in vivo validation, and highlighting its potential impact on the development of new anticancer therapies.

A Novel Deep Transfer Learning-Based Approach for Face Pose Estimation

Abstract An efficient face recognition system is essential for security and authentication-based applications. However, real-time face recognition systems have a few significant concerns, including face pose orientations. In the last decade, numerous solutions have been introduced to estimate distinct face pose orientations. Nevertheless, these solutions must be adequately addressed for the three main face pose orientations: Yaw, Pitch, and Roll. This paper proposed a novel deep transfer learning-based multitasking approach for solving three integrated tasks, i.e., face detection, landmarks detection, and face pose estimation. The face pose variation vulnerability has been intensely investigated here underlying three modules: image preprocessing, feature extraction module through deep transfer learning, and regression module for estimating the face poses. The experiments are performed on the well-known benchmark dataset Annotated Faces in the Wild (AFW). We evaluate the outcomes of the experiments to reveal that our proposed approach is superior to other recently available solutions.

NO CHEEK BIAS FOR NON-PRIMATES: AN INSTAGRAM REPLICATION OF THOMAS ET AL. (2006)

Previous research has established that photos of great apes, including humans, show a left cheek bias. As this bias is absent in images of lower primates and other animals, phylo-genetic proximity appears to influence humans’ depictions of nonhuman species. However Thomas et al.’s (2006) finding of a left cheek bias for dogs challenges this argument. As their analyses were underpowered, the present study sought to replicate Thomas et al.’s study with a larger sample to help determine whether human depictions of non-human animals vary as a function of their evolutionary relatedness.Photographs (N=2883) were sourced from Instagram’s ‘Most Recent’ feed using hashtags that matched Thomas et al.’s Google Image search terms: #dog, #cat, #fish, #lizard, #cute- baby, #cryingbaby. The first 401 lateral images for each hashtag were coded for pose orientation (left, right).Replicating Thomas et al., results confirmed a left cheek bias for mammals but not nonmammals. The left cheek bias was driven by images of human infants; there were no cheek biases for images of nonhuman animals (dogs, cats, lizards, fish).As a left cheek bias was evident in photos of primates (#cutebaby, #cryingbaby), but absent for other mammals (#dog, #cat) and nonmammals (#lizard, #fish), the data support the argument that phylogenetic proximity influences posing biases.

Head pose estimation: An extensive survey on recent techniques and applications

Biometric based systems are involved in many areas, from surveillance to user authentication, from autonomous systems to human-robot interactions. Head pose estimation (HPE) is the task to support biometric systems in which any of the biometric traits of the head is involved, as face, ear or iris. This particular biometric branch finds its application in driver attention detection, surveillance for recognition, face frontalization, best frame selection and so on. The goal of HPE is to determine the head pose orientation (yaw, pitch, roll). The implemented methods use different techniques depending on the kind of input. In this survey we present an overview of involved datasets, recent techniques and applications. We evaluate and compare the different approaches with respect to their advantages and practical usage. In addition, we propose a technical comparison between training and training-free techniques for the most popular HPE methods.

A Video Analytic In-Class Student Concentration Monitoring System

Automatic learning feedback monitoring and analysis are becoming essential in modern education. We present a video analytic system capable of monitoring in-class student’s learning behaviors and providing feedback to the instructor. It is a common practice nowadays for students to take electronic notes or browse online using laptops and cellphones in class. However the use of technology can also impact student concentration and affect learning behaviors, which can seriously hinder their learning progress if not controlled properly. In this pioneering study, we propose a non-intrusive deep-learning based computer vision system to monitor student concentration by extracting and inferring high-level visual behavior cues, including their facial expressions, gestures and activities. Our system can automatically assist instructors with situational awareness in real time. We assume only RGB color images as input and runable system on edge devices for easy deployment. We propose two video analytic components for student behavior analysis: (1) The facial analysis component operates based on Dlib face detection and facial landmark tracking to localize each student and analyze their face orientations, eye blinking, gazes, and facial expressions. (2) The activity detection and recognition component operates based on OpenPose and COCO object detection can identify eight types of in-class gestures and behaviors including raising-hand, typing, phone-answering, crooked-head, desk napping, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">etc.</i> Experiments are performed on a newly collected real-world In-Class Student Activity Dataset (ICSAD), where we achieved nearly 80% activity detection rate. Our system is view-independent in handling facial and pose orientations with average angular error < 10°. The source code of this work is at: <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/YiZengHsieh/ICSAD</uri> .

Immunoproteasome and Non-Covalent Inhibition: Exploration by Advanced Molecular Dynamics and Docking Methods.

The selective inhibition of immunoproteasome is a valuable strategy to treat autoimmune, inflammatory diseases, and hematologic malignancies. Recently, a new series of amide derivatives as non-covalent inhibitors of the β1i subunit with Ki values in the low/submicromolar ranges have been identified. Here, we investigated the binding mechanism of the most potent and selective inhibitor, N-benzyl-2-(2-oxopyridin-1(2H)-yl)propanamide (1), to elucidate the steps from the ligand entrance into the binding pocket to the ligand-induced conformational changes. We carried out a total of 400 ns of MD-binding analyses, followed by 200 ns of plain MD. The trajectories clustering allowed identifying three representative poses evidencing new key interactions with Phe31 and Lys33 together in a flipped orientation of a representative pose. Further, Binding Pose MetaDynamics (BPMD) studies were performed to evaluate the binding stability, comparing 1 with four other inhibitors of the β1i subunit: N-benzyl-2-(2-oxopyridin-1(2H)-yl)acetamide (2), N-cyclohexyl-3-(2-oxopyridin-1(2H)-yl)propenamide (3), N-butyl-3-(2-oxopyridin-1(2H)-yl)propanamide (4), and (S)-2-(2-oxopyridin-1(2H)-yl)-N,4-diphenylbutanamide (5). The obtained results in terms of free binding energy were consistent with the experimental values of inhibition, confirming 1 as a lead compound of this series. The adopted methods provided a full dynamic description of the binding events, and the information obtained could be exploited for the rational design of new and more active inhibitors.

RCBi-CenterNet: An Absolute Pose Policy for 3D Object Detection in Autonomous Driving

3D Object detection is a critical mission of the perception system of a self-driving vehicle. Existing bounding box-based methods are hard to train due to the need to remove duplicated detections in the post-processing stage. In this paper, we propose a center point-based deep neural network (DNN) architecture named RCBi-CenterNet that predicts the absolute pose for each detected object in the 3D world space. RCBi-CenterNet is composed of a recursive composite network with a dual-backbone feature extractor and a bi-directional feature pyramid network (BiFPN) for cross-scale feature fusion. In the detection head, we predict a confidence heatmap that is used to determine the position of detected objects. The other pose information, including depth and orientation, is regressed. We conducted extensive experiments on the Peking University/Baidu-Autonomous Driving dataset, which contains more than 60,000 labeled 3D vehicle instances from 5277 real-world images, and each vehicle object is annotated with the absolute pose described by the six degrees of freedom (6DOF). We validated the design choices of various data augmentation methods and the backbone options. Through an ablation study and an overall comparison with the state-of-the-art (SOTA), namely CenterNet, we showed that the proposed RCBi-CenterNet presents performance gains of 2.16%, 2.76%, and 5.24% in Top 1, Top 3, and Top 10 mean average precision (mAP). The model and the result could serve as a credible benchmark for future research in center point-based object detection.

Robust grasp detection with incomplete point cloud and complex background

Point cloud information is a convenient means to accomplish grasp detection in autonomous robotic grasping. However, the performance of various state-of-the-art grasp detection methods that use point cloud decreases significantly when the available point cloud information is incomplete and the background is complex, i.e. heterogeneous. To solve this problem, we propose a robust grasp detection method that demonstrates higher performance, especially with incomplete point clouds and complex backgrounds. We introduce a novel technique named ‘visible point-cloud’ – generated using the point cloud and pose (position and orientation) information of the sensor(s) – that helps to eliminate unsafe grasp candidates quickly and efficiently. The remaining grasp candidates are then classified and the best candidate is determined using a cost function. The effectiveness of the proposed method is shown experimentally using a 6-DoF robot arm equipped with a two-finger gripper with three different background settings: (a) steps, (b) pillars, and (c) table top. The results show that the proposed method is 1.20 times faster and has a 20% higher grasp success rate than a state-of-the-art method for a single point cloud camera. The results demonstrate that the proposed method significantly improves the performance of autonomous grasping even with incomplete point cloud and is robust for different backgrounds.

Radial Coverage Strength for Optimization of Monocular Multicamera Deployment

In this article, a new concept, radial coverage strength, is first proposed to characterize the visual sensing performance when the orientation of the target pose is considered. In particular, the elevation angle of the optical pose of the visual sensor is taken to decompose the visual coverage strength into effective and ineffective components, motivated by the imaging intuition. An optimization problem is then formulated for a monocular multicamera network to maximize the coverage of the object area based on the strength information fusion along the effective coverage strength direction through the deployment of the angle between radial coverage vector of the camera optical pose. Both simulation and experiments are conducted to validate the proposed approach and comparison with existing methods is also provided.

Phylogenetic proximity influences humans’ depictions of nonhuman primates on Instagram

ABSTRACT Portraits of humans favour the left cheek, with emotion thought to drive this posing asymmetry. In primates the emotion-dominant right hemisphere predominantly controls the left hemiface, rendering the left cheek anatomically more expressive than the right. As perceptions of nonhuman primates vary with genetic relatedness, depictions of nonhuman primates should theoretically be influenced by their phylogenetic proximity to humans. The present study thus examined whether humans depict nonhuman primates showing the left cheek, and whether depictions vary with evolutionary distance. Photographs of nonhuman primates were sourced from Instagram’s “Most recent” feed: great apes (#chimpanzee, #bonobo, #gorilla, #orangutan), lesser apes (#gibbon), Old World monkeys (#baboon, #macaque, #proboscismonkey), New World monkeys (#spidermonkey, #marmosetmonkey, #capuchin), and prosimians (#lemur, #slowloris, #tarsier). The first 500 single-subject images for each hashtag (except #slowloris for which 318 images were available) were coded for pose orientation (left, right) and portrait type (head/torso, full body). As anticipated, there was a left cheek bias for great apes but no bias for more distantly related primates. These data thus suggest that depictions of nonhuman primates are implicitly influenced by phylogenetic proximity: the more closely related the primate, the more likely we are to depict them as we do ourselves, showing the left cheek.

State Estimation and Localization Based on Sensor Fusion for Autonomous Robots in Indoor Environment

Currently, almost all robot state estimation and localization systems are based on the Kalman filter (KF) and its derived methods, in particular the unscented Kalman filter (UKF). When applying the UKF alone, the estimate of the state is not sufficiently precise. In this paper, a new hierarchical infrared navigational algorithm hybridization (HIRNAH) system is developed to provide better state estimation and localization for mobile robots. Two navigation subsystems (inertial navigation system (INS) and, using a novel infrared navigation algorithm (NIRNA), Odom-NIRNA) and an RPLIDAR-A3 scanner cooperation to build HIRNAH. The robot pose (position and orientation) errors are estimated by a system filtering module (SFM) and used to smooth the robot’s final poses. A prototype (two rotary encoders, one smartphone-based robot sensing model and one RPLIDAR-A3 scanner) has been built and mounted on a four-wheeled mobile robot (4-WMR). Simulation results have motivated real-life experiments, and obtained results are compared to some existent research (hardware and control technology navigation (HCTNav), rapid exploring random tree (RRT) and in stand-alone mode (INS)) for performance measurements. The experimental results confirm that HIRNAH presents a more accurate estimation and a lower mean square error (MSE) of the robot’s state than those calculated by the previously cited HCTNav, RRT and INS.

Face recognition using machine vision for Security system

The aim of this project is to design and develop a face recognition based security system using machine vision. In this paper, a face recognition system using Principal Component Analysis (PCA) with Euclidean distance classifier is proposed. PCA is selected as it is less sensitive to noise and interference, besides reducing the number of variables in face recognition which reduce computational time and increase accuracy. A number of experiments were done to evaluate the performance of the face recognition system. A training database of 20 students is created and used in this project. Overall, the face recognition system on static face images has 86% success rate and 14% error rate which is mainly due to the variation of pose orientation. It is also found that the threshold value is an important factor for the performance of face recognition. The distance of the person to the camera play a role as the further the person is away from the camera, the more blur the face image captured. As for the anti-spoofing approach, the proposed detection of eye blinking is successfully tested with training eyes open and eyes close images.

Humans’ left cheek portrait bias extends to chimpanzees: Depictions of chimps on Instagram

ABSTRACT When posing for portraits, humans favour the left cheek. This preference is argued to stem from the left cheek’s greater expressivity: as the left hemiface is predominantly controlled by the emotion-dominant right hemisphere, it expresses emotion more intensely than the right hemiface. Whether this left cheek bias extends to our closest primate relatives, chimpanzees, has yet to be determined. Given that humans and chimpanzees share the same oro-facial musculature and contralateral cortical innervation of the face, it appears probable that humans would also choose to depict chimps showing the more emotional left cheek. This paper thus examined portrait biases in images of chimpanzees. Two thousand photographs were sourced from Instagram’s “Most Recent” feed using the #chimpanzee, and coded for pose orientation (left, right) and portrait type (head and torso, full body). As anticipated, there were significantly more left cheek (57.2%) than right cheek images (42.8%), with the bias observed across both head and torso and full body portraits. Thus humans choose to depict chimpanzees just as we depict ourselves: offering the left cheek. As such, these data confirm that the left cheek bias is observed in both human and non-human primates, consistent with an emotion-based account of the orientation preference.

Estimation of Pedestrian Pose Orientation Using Soft Target Training Based on Teacher⁻Student Framework.

Semi-supervised learning is known to achieve better generalisation than a model learned solely from labelled data. Therefore, we propose a new method for estimating a pedestrian pose orientation using a soft-target method, which is a type of semi-supervised learning method. Because a convolutional neural network (CNN) based pose orientation estimation requires large numbers of parameters and operations, we apply the teacher–student algorithm to generate a compressed student model with high accuracy and compactness resembling that of the teacher model by combining a deep network with a random forest. After the teacher model is generated using hard target data, the softened outputs (soft-target data) of the teacher model are used for training the student model. Moreover, the orientation of the pedestrian has specific shape patterns, and a wavelet transform is applied to the input image as a pre-processing step owing to its good spatial frequency localisation property and the ability to preserve both the spatial information and gradient information of an image. For a benchmark dataset considering real driving situations based on a single camera, we used the TUD and KITTI datasets. We applied the proposed algorithm to various driving images in the datasets, and the results indicate that its classification performance with regard to the pose orientation is better than that of other state-of-the-art methods based on a CNN. In addition, the computational speed of the proposed student model is faster than that of other deep CNNs owing to the shorter model structure with a smaller number of parameters.

Pose Consensus of Multiple Robots with Time-Delays Using Neural Networks

SummaryThis paper proposes a novel control scheme based on Radial Basis Artificial Neural Network to solve the leader–follower and leaderless pose (position and orientation) consensus problems in the Special Euclidean space of dimension three (SE(3)). The controller is designed for robot networks composed of heterogeneous (kinematically and dynamically different) and uncertain robots with variable time-delays in the interconnection. The paper derives a sufficient condition on the controller gains and the robot interconnection, and using Barbalat’s Lemma, both consensus problems are solved. The proposed approach employs the singularity-free, unit-quaternions to represent the orientation of the end-effectors in theSE(3). The significance and advantages of the proposed control scheme are that it solves the two pose consensus problems for heterogeneous robot networks considering variable time-delays in the interconnection without orientation representation singularities, and the controller does not require to know the dynamic model of the robots. The performance of the proposed controller is illustrated via simulations with a heterogeneous robot network composed of robots with 6-DoF and 7-DoF.

Left cheek poses garner more likes: the effect of pose orientation on Instagram engagement

ABSTRACTIn social media’s attention economy “likes” are currency; photos showing faces attract more “likes.” Previous research has established a left cheek bias in photos uploaded to social media, but whether left cheek poses induce more engagement than right cheek poses remains to be determined. The present study thus examined whether pose orientation influences the number of “likes” and comments garnered by photos uploaded to Instagram. The top 20 single-user Instagram accounts were identified, and the most recent 10 left and 10 right cheek images were selected, resulting in a total of 400 images. The number of “likes” and comments were tallied for each image, netting over 1 billion “likes” and 14 million comments for analysis. Results confirmed that pose orientation influences audience engagement: left cheek poses garner >10% more “likes” than right cheek poses. Gender did not influence “likes”. Comments were not affected by either pose orientation or gender, likely reflecting the different levels of effort and motivations involved in “liking” vs. commenting on an image. These data indicate that a seemingly inconsequential turn of the head profoundly impacts audience engagement: left cheek poses gained >330,000 more “likes”, offering clear implications for marketers and others in the social media economy.

Observability and sensitivity analysis of lightcurve measurement models for use in space situational awareness

ABSTRACTThis paper deals with the observability and sensitivity analysis of the lightcurve measurement for use in the estimation of instantaneous pose (orientation) and shape geometry. Since several reflectance models exist that are typically used in obtaining a synthetic lightcurve measurement, they are compared to assess the observability or/and ‘sensitivity’ to determine the ‘best model’ for use. These measurement models are nonlinear with implicit and non-trivial dependency on the states; a numerical Jacobian as well as an unscented Kalman filter derived observation matrix are synthesized for each choice of the measurement model and compared. As the linearization is about an attitude (orientation) trajectory, distinct cases are considered to elaborate the results. For the cases evaluated here, it is shown that the attitude and shape/size can indeed be estimated from the lightcurve, but this is also dependent upon the initial conditions (subsequent attitude trajectory). Linear observability analysis of a discretized system is also performed with respect to the attitude states and shape/size parameters using the singular value decomposition of the observability matrix synthesized for a batch of measurements. The results are summarized for various initial conditions of the resident space object's attitude and angular velocity states.

A Dual Quaternion Feedback Linearized Approach for Maneuver Regulation of Rigid Bodies

The adoption of the dual quaternion formalism to represent the pose (position and orientation) of a rigid body allows to design a single controller to regulate both its position and its attitude. In this letter, we adopt such a pose representation to develop an exponentially stable maneuver regulation control law, ensuring robust path following in the presence of disturbances. The designed solution relies on the feedback linearized model of the dual quaternion based dynamics of the rigid body. Numerical results confirm the effectiveness of the proposed maneuver regulation approach when compared with trajectory tracking in a noisy scenario.

On the Importance of Uncertainty Representation in Active SLAM

The purpose of this work is to highlight the paramount importance of representing and quantifying uncertainty to correctly report the associated confidence of the robot's location estimate at each time step along its trajectory and therefore decide the correct course of action in an active SLAM mission. We analyze the monotonicity property of different decision-making criteria, both in 2-D and 3-D, with respect to the representation of uncertainty and of the orientation of the robot's pose. Monotonicity, the property that uncertainty increases as the robot moves, is essential for adequate decision making. We analytically show that, by using differential representations to propagate spatial uncertainties, monotonicity is preserved for all optimality criteria, A-opt , D-opt , and E-opt , and for Shannon's entropy. We also show that monotonicity does not hold for any criteria in absolute representations using Roll-Pitch-Yaw and Euler angles. Finally, using unit quaternions in absolute representations, the only criteria that preserve monotonicity are D-opt and Shannon's entropy.

Joining the dots: reconstructing 3D environments and movement paths using animal-borne devices

BackgroundAnimal-attached sensors are increasingly used to provide insights on behaviour and physiology. However, such tags usually lack information on the structure of the surrounding environment from the perspective of a study animal and thus may be unable to identify potentially important drivers of behaviour. Recent advances in robotics and computer vision have led to the availability of integrated depth-sensing and motion-tracking mobile devices. These enable the construction of detailed 3D models of an environment within which motion can be tracked without reliance on GPS. The potential of such techniques has yet to be explored in the field of animal biotelemetry. This report trials an animal-attached structured light depth-sensing and visual–inertial odometry motion-tracking device in an outdoor environment (coniferous forest) using the domestic dog (Canis familiaris) as a compliant test species.ResultsA 3D model of the forest environment surrounding the subject animal was successfully constructed using point clouds. The forest floor was labelled using a progressive morphological filter. Trees trunks were modelled as cylinders and identified by random sample consensus. The predicted and actual presence of trees matched closely, with an object-level accuracy of 93.3%. Individual points were labelled as belonging to tree trunks with a precision, recall, and F_{beta } score of 1.00, 0.88, and 0.93, respectively. In addition, ground-truth tree trunk radius measurements were not significantly different from random sample consensus model coefficient-derived values. A first-person view of the 3D model was created, illustrating the coupling of both animal movement and environment reconstruction.ConclusionsUsing data collected from an animal-borne device, the present study demonstrates how terrain and objects (in this case, tree trunks) surrounding a subject can be identified by model segmentation. The device pose (position and orientation) also enabled recreation of the animal’s movement path within the 3D model. Although some challenges such as device form factor, validation in a wider range of environments, and direct sunlight interference remain before routine field deployment can take place, animal-borne depth sensing and visual–inertial odometry have great potential as visual biologging techniques to provide new insights on how terrestrial animals interact with their environments.