2D Projection Research Articles

Determining the articulatory parameters in the "Avantis 3D" software, taking into account the position of the individual hinge axis.

In modern dentistry, the tactics of treating patients with occlusion and TMJ pathology are based on complex diagnostics. The key factor for the planned treatment of a patient depends on the accuracy of recording the individual parameters of the movement of the lower jaw. One of the methods for diagnosing the work of the temporomandibular joint and obtaining individual parameters for setting up the articulator is axiography (condylography). With the introduction of digital technologies into the practice of a dentist, such as CBCT, intraoral scanning, and the necessary software for creating 3D projects, the well-known method of registering protrusion and laterotrusion for setting the angles of the sagittal articular path and the angles of the Bennett joint angles in the articulator, based on the Christensen phenomenon, has found a new application. The creation of a virtual twin of the patient, the determination of his individual articulatory parameters greatly simplifies the procedure and reduces the time of diagnosis for the subsequent full rehabilitation of a patient with TMJ dysfunction.

A Hybrid Recognition Framework for Highly Interacting Machining Features Based on Primitive Decomposition, Learning and Reconstruction

For the highly interacting machining features, Layered Projection Decomposition Method presents inferior recognition efficiency and accuracy, due to its high-cost 3D projection and failures in determining projection faces for internal occluded faces. To address these issues, we propose a potential hybrid recognition framework. We first introduce a straightforward adjacent projection wire (APW) over UV wires, automatically restoring the full projection wires from highly interacting features. Building on APWs, an efficient hybrid boundary representation and its corresponding unambiguous primitive definitions are proposed by combing with graph-based boundary representations. Subsequently, we design an efficient primitive decomposition method by introducing primitive boundary matching to decide the initial projection faces, and introducing iterative projection boundary expansion to complete the full primitives from occluded faces. Moreover, we establish an efficient Graph Neural Network to learn the distinguishable distributions over the decomposed primitives. Specifically, an Adjacency Attention Unit is proposed to automatically perceive the influence weight of adjacent nodes, leading to more discriminative self-adaptive shape embedding for efficient primitive recognition. Finally, we summarize convenient reconstruction rules to correct the wrong predictions of feature faces with indistinguishable adjacent relationships. To evaluate the effectiveness of the proposed recognition framework, CAD models of complex aircraft structural parts are collected to present a challenging machining feature dataset. Extensive numerical experiments demonstrate that the proposed hybrid recognition framework enables significant improvements over the state-of-the-art machining feature recognition techniques.

Kinematical signatures: Distinguishing between warps and radial flows

Increasing evidence shows that warped disks are common, challenging the methods used to model their velocity fields. Molecular line emission of these disks is characterized by a twisted pattern, similar to the signal from radial flows, complicating the study of warped disk kinematics. Previous attempts to model these features have encountered difficulties in distinguishing between the underlying kinematics of different disks. This study aims to advance gas kinematics modeling capabilities by extending the Extracting Disk Dynamics ( eddy ) package to include warped geometries and radial flows. We assess the performance of eddy in recovering input parameters for scenarios involving warps, radial flows, and combinations of the two. Additionally, we provide a basis to break the visual degeneracy between warped disks and radial flow, establishing a criterion to distinguish them. We extended the eddy package to handle warped geometries by including a parametric prescription of a warped disk and a ray-casting algorithm to account for the surface self-obscuration arising from the 3D to 2D projection. The effectiveness of the tool was tested using the radiative transfer code RADMC3D generating synthetic models for disks with radial flows, warped disks, and warped disks with radial flows. We demonstrate the efficacy of our tool in accurately recovering the geometrical parameters of systems, particularly in data with sufficient angular resolution. Importantly, we observe minimal impact from thermal noise levels typical in Atacama Large Millimeter/submillimeter Array (ALMA) observations. Furthermore, our findings reveal that fitting an incorrect model type produces characteristic residual signatures, which serve as kinematic criteria for disk classification. Characterizing gas kinematics requires careful consideration of twisted motions. While our model provides insights into disk geometries, caution is needed when interpreting parameters in regions with complex kinematics or low-resolution data. Future ALMA baseline observations should help clarify warped disk kinematics.

New Solution to 3D Projection in Human-like Binocular Vision

A human eye has about 120 million rod cells and 6 million cone cells. This huge number of light-sensing cells inside a human eye will continuously produce a huge quantity of visual signals that flow into the human brain for daily processing. However, the real-time processing of these visual signals does not cause excessive energy consumption by the human brain. This fact tells us the truth which is to say that human-like vision processes do not rely on complicated and expensive formulas to compute depth, displacement, and colors. On the other hand, the human eye is like a camera with pan-tilt (PT) motions. We all know that in computer vision, each set of PT parameters (i.e., coefficients of pan motion as well as tilt motion) requires a dedicated calibration to determine a camera’s projection matrix. Since there is an infinite number of PT parameters that could be produced by a human eye, it is unlikely that a human brain stores an infinite number of calibration matrices for each human eye. These observations inspire us to look for a simpler and computationally non-expensive solution which is to undertake three-dimensional (3D) projection in human-like binocular vision. In other words, it is an interesting question for us to answer, which is to say whether simpler and learning-friendly formulas for computing depth and displacement exist or not. If the answer is yes, these formulas must also be calibration-friendly (i.e., easy process on the fly or on the go). In this paper, we present an important discovery of a new solution to 3D projection in a human-like binocular vision system. This solution is computationally simpler and could be easily learned or calibrated on the fly. We know that the purpose of doing 3D projection in binocular vision is to undertake forward and inverse transformations (or mappings) between coordinates in 2D digital images and coordinates in a 3D analogue scene. The formulas underlying the new solution are accurate, easily computable, easily tunable (i.e., to be calibrated on the fly or on the go) and could be easily implemented by a neural system (i.e., a network of neurons or a network of computational flows). Experimental results have validated the newly derived formulas which are better than textbook solutions.

First-principles calculations of the elastic anisotropy, thermal conductivity, and optical properties of MAX-phase M2AlC (M = V, Nb, Ta) series carbides

Materials from the MAX phase, such as the M2AlC (M = V, Nb, Ta) series carbides, exhibit superior mechanical, electronic, and thermodynamic properties, making them highly promising for research. Density functional theory (DFT) has been employed to study the elastic anisotropy, thermal conductivity, electronic, and optical properties of these carbides. The results indicate that the formation enthalpies of the M2AlC (M = V, Nb, Ta) series are negative and the phonon spectra do not exhibit imaginary frequencies, confirming the thermodynamic and dynamic stability of these carbides. Additionally, the calculated Poisson's ratios and B/G ratios suggest that these materials are brittle and contain covalent and ionic bonds—a conclusion further corroborated by analyses of electronic structures. The elastic anisotropy is illustrated through anisotropy indices, three-dimensional (3D) surface configurations, and two-dimensional (2D) projections, with the sequence of anisotropy being: Nb2AlC > Ta2AlC > V2AlC. Further investigations into the sound velocities, Debye temperature, anisotropy of sound velocity, thermal conductivity, and optical properties of the M2AlC (M = V, Nb, Ta) series provide substantial theoretical support for the research on MAX phase M2AlC (M = V, Nb, Ta) series carbides.

Fast reconstruction 3D computed tomography image of stacked cell under faster scanning by dual-branch cross-fusion flat bottom network

Abstract Cone beam computed tomography (CBCT) fast scanning and reconstruction is a key step to achieve rapid detection of internal defects in batteries. In this work, we have achieved a faster CT scanning just in 5 seconds by reducing the X-ray exposure time in sparse view CT. However, the CT data is extremely incomplete by faster scanning; the existing reconstruction methods are difficult to reconstruct a high quality three-dimensional (3D) CT image of stacked cells. To address this issue, we propose a 3D CT image reconstruction network, which can reconstruct higher quality CT images from low quality 3D volume data. The input data of the reconstruction network is not 2D projection data, but 3D volume data. In this network, a high and low resolution dual-branch cross-fusion flat bottom structure is designed. The high resolution flat bottom branch aims to preserve detailed information, while the low resolution flat bottom branch focuses on capturing more semantic information. Cross-fusion between these branches mitigates the loss of semantic details. Additionally, the auxiliary loss function, the main loss function, and the 3D attention module are designed to enhance semantic accuracy and the learning performance of the network. The 3D training data is collected under a fast scanning strategy spanning 5-60 seconds. During the training phase, we use clipping block technology to cut the 3D volume data, enabling direct training on the 3D volume data. Our experimental results demonstrate that our 3D reconstruction network outperforms mainstream algorithms under this faster scanning strategy, which is able to reconstruct higher quality 3D CT images just in 15 seconds. Ablation experiments confirm the positive impact of the dual-branch cross-fusion flat bottom structure, attention module, and loss functions on improving the quality of 3D CT images.

Deep learning architecture for scatter estimation in cone-beam computed tomography head imaging with varying field-of-measurement settings.

X-ray scatter causes considerable degradation in the cone-beam computed tomography (CBCT) image quality. To estimate the scatter, deep learning-based methods have been demonstrated to be effective. Modern CBCT systems can scan a wide range of field-of-measurement (FOM) sizes. Variations in the size of FOM can cause a major shift in the scatter-to-primary ratio in CBCT. However, the scatter estimation performance of deep learning networks has not been extensively evaluated under varying FOMs. Therefore, we train the state-of-the-art scatter estimation neural networks for varying FOMs and develop a method to utilize FOM size information to improve performance. We used FOM size information as additional features by converting it into two channels and then concatenating it to the encoder of the networks. We compared our approach for a U-Net, Spline-Net, and DSE-Net, by training them with and without the FOM information. We utilized a Monte Carlo-simulated dataset to train the networks on 18 FOM sizes and test on 30 unseen FOM sizes. In addition, we evaluated the models on the water phantoms and real clinical CBCT scans. The simulation study demonstrates that our method reduced average mean-absolute-percentage-error for U-Net by 38%, Spline-Net by 40%, and DSE-net by 33% for the scatter estimation in the 2D projection domain. Furthermore, the root-mean-square error on the 3D reconstructed volumes was improved for U-Net by 43%, Spline-Net by 30%, and DSE-Net by 23%. Furthermore, our method improved contrast and image quality on real datasets such as water phantom and clinical data. Providing additional information about FOM size improves the robustness of the neural networks for scatter estimation. Our approach is not limited to utilizing only FOM size information; more variables such as tube voltage, scanning geometry, and patient size can be added to improve the robustness of a single network.

Accuracy of fish-eye camera device for aircraft positioning based on laboratory experiment combining computer graphic and 3D Projection

For the management of airborne mobility devices such as airplanes and UAVs, such as noise assessment, it is important to measure the three-dimensional position of the aircraft, and various technologies have been researched and developed. We have developed an Aircraft Positioning Camera (APC) based on an IoT device equipped with a fisheye camera that can mechanically measure aircraft position at a local area with low cost and with high mobility. The performance of this APC is still uncertain as to the limits of the flight altitude and speed of the aircraft that can be measured. Therefore, a detailed study of its performance is needed. Since, it is difficult to standardize the conditions such as weather and ground surface conditions, it is effective to conduct these studies in a laboratory where environmental changes can be controlled. The performance of this APC was examined at an experimental facility where images captured by a 360-degree camera or computer graphics can be projected in three dimensions onto a hemispherical dome.

3D human model guided pose transfer via progressive flow prediction network

Human pose transfer is to transfer a conditional person image to a new target pose. The difficulty lies in modeling the large-scale spatial deformation from the conditional pose to the target one. However, the commonly used 2D data representations and one-step flow prediction scheme lead to unreliable deformation prediction because of the lack of 3D information guidance and the great changes in the pose transfer. Therefore, to bring the original 3D motion information into human pose transfer, we propose to simulate the generation process of real person image. We drive the 3D human model reconstructed from the conditional person image with the target pose and project it to the 2D plane. The 2D projection thereby inherits the 3D information of the poses which can guide the flow prediction. Furthermore, we propose a progressive flow prediction network consisting of two streams. One stream is to predict the flow by decomposing the complex pose transformation into multiple sub-transformations. The other is to generate the features of the target image according to the predicted flow. Besides, to enhance the reliability of the generated invisible regions, we use the target pose information which contains structural information from the flow prediction stream as the supplementary information to the feature generation. The synthesized images with accurate depth information and sharp details demonstrate the effectiveness of the proposed method.

A Framework for Assessing Ocean Mixed Layer Depth Evolution

AbstractThe ocean surface mixed layer plays a crucial role as an entry or exit point for heat, salt, momentum, and nutrients from the surface to the deep ocean. In this study, we introduce a framework to assess the evolution of the mixed layer depth (MLD) for realistic forcings and preconditioning conditions. Our approach involves a physically‐based parameter space defined by three dimensionless numbers: λs representing the relative contribution of the buoyancy flux and the wind stress at the air‐sea interface, Rh the Richardson number which characterizes the stability of the water column relative to the wind shear, and f/Nh which characterizes the importance of the Earth's rotation (ratio of the Coriolis frequency f and the pycnocline stratification Nh). Four MLD evolution regimes (“restratification,” “stable,” “deepening,” and “strong deepening”) are defined based on the values of the normalized temporal evolution of the MLD. We evaluate the 3D parameter space in the context of 1D simulations and we find that considering only the two dimensions (λs, Rh) is the best choice of 2D projection of this 3D parameter space. We then demonstrate the utility of this two‐dimensional λs − Rh parameter space to compare 3D realistic ocean simulations: we discuss the impact of the horizontal resolution (1°, 1/12°, or 1/60°) and the Gent‐McWilliams parameterization on MLD evolution regimes. Finally, a proof of concept of using observational data as a truth indicates how the parameter space could be used for model calibration.

Large-Area Perovskite Nanocrystal Metasurfaces for Direction-Tunable Lasing.

Perovskite nanocrystals (PNCs) are attractive emissive materials for developing compact lasers. However, manipulation of PNC laser directionality has been difficult, which limits their usage in photonic devices that require on-demand tunability. Here we demonstrate PNC metasurface lasers with engineered emission angles. We fabricated millimeter-scale CsPbBr3 PNC metasurfaces using an all-solution-processing technique based on soft nanoimprinting lithography. By designing band-edge photonic modes at the high-symmetry X point of the reciprocal lattice, we achieved four linearly polarized lasing beams along a polar angle of ∼30° under optical pumping. The device architecture further allows tuning of the lasing emission angles to 0° and ∼50°, respectively, by adjusting the PNC thickness to shift other high-symmetry points (Γ and M) to the PNC emission wavelength range. Our laser design strategies offer prospects for applications in directional optical antennas and detectors, 3D laser projection displays, and multichannel visible light communication.

Controlling the scatterplot shapes of 2D and 3D multidimensional projections

Multidimensional projections are effective techniques for depicting high-dimensional data. The point patterns created by such techniques, or a technique’s visual signature, depend — apart from the data themselves — on the technique design and its parameter settings. Controlling such visual signatures — something that only few projections allow — can bring additional freedom for generating insightful depictions of the data. We present a novel projection technique — ShaRP — that allows explicit control on such visual signatures in terms of shapes of similar-value point clusters (settable to rectangles, triangles, ellipses, and convex polygons) and the projection space (2D or 3D Euclidean or S2). We show that ShaRP scales computationally well with dimensionality and dataset size, provides its signature-control by a small set of parameters, allows trading off projection quality to signature enforcement, and can be used to generate decision maps to explore the behavior of trained machine-learning classifiers.

A deep learning method for simultaneous denoising and missing wedge reconstruction in cryogenic electron tomography

Cryogenic electron tomography is a technique for imaging biological samples in 3D. A microscope collects a series of 2D projections of the sample, and the goal is to reconstruct the 3D density of the sample called the tomogram. Reconstruction is difficult as the 2D projections are noisy and can not be recorded from all directions, resulting in a missing wedge of information. Tomograms conventionally reconstructed with filtered back-projection suffer from noise and strong artefacts due to the missing wedge. Here, we propose a deep-learning approach for simultaneous denoising and missing wedge reconstruction called DeepDeWedge. The algorithm requires no ground truth data and is based on fitting a neural network to the 2D projections using a self-supervised loss. DeepDeWedge is simpler than current state-of-the-art approaches for denoising and missing wedge reconstruction, performs competitively and produces more denoised tomograms with higher overall contrast.

Multi-Grained Point Cloud Geometry Compression via Dual-Model Prediction with Extended Octree

The state-of-the-art geometry-based point cloud compression (G-PCC) (Octree) is the fine-grained approach, which uses the octree to partition point clouds into voxels and predicts them based on neighbor occupancy in narrower spaces. However, G-PCC (Octree) is less effective at compressing dense point clouds than multi-grained approaches (such as G-PCC (Trisoup)), which exploit the continuous point distribution in nodes partitioned by the pruned octree over larger spaces. Therefore, we propose a lossy multi-grained compression with extended octree and dual-model prediction. The extended octree, where each partitioned node contains intra-block and extra-block points, is applied to address poor prediction (such as overfitting) at the node edges of the octree partition. For the points of each multi-grained node, dual-model prediction fits surfaces and projects residuals onto the surfaces, reducing projection residuals for efficient 2D compression and fitting complexity. In addition, a hybrid DWT-DCT transform for 2D projection residuals mitigates the resolution degradation of DWT and the blocking effect of DCT during high compression. Experimental results demonstrate the superior performance of our method over advanced G-PCC (Octree), achieving BD-Rate gains of 55.9% and 45.3% for point-to-point ( D1 ) and point-to-plane ( D2 ) distortions, respectively. Our approach also outperforms G-PCC (Octree) and G-PCC (Trisoup) in subjective evaluation.

Three-dimensional magnetization reconstruction from electron optical phase images with physical constraints

The ability to characterize three-dimensional (3D) magnetization distributions in nanoscale magnetic materials and devices is essential to fully understand their static and dynamic magnetic properties. Phase contrast techniques in the transmission electron microscope (TEM), such as electron holography and electron ptychography, can be used to record two-dimensional (2D) projections of the in-plane magnetic induction of 3D nanoscale objects. Although the 3D magnetic induction can in principle be reconstructed from one or more tilt series of such 2D projections, conventional tomographic reconstruction algorithms do not recover the 3D magnetization within a sample directly. Here, we use simulations to describe the basis of an improved model-based algorithm for the tomographic reconstruction of a 3D magnetization distribution from one or more tilt series of electron optical phase images recorded in the TEM. The algorithm allows a wide range of physical constraints, including a priori information about the sample geometry and magnetic parameters, to be specified. It also makes use of minimization of the micromagnetic energy in the loss function. We demonstrate the reconstruction of the 3D magnetization of a localized magnetic soliton — a hopfion ring — and discuss the influence of noise, choice of magnetic constants, maximum tilt angle and number of tilt axes on the result. The algorithm can in principle be adapted for other magnetic contrast imaging techniques in the TEM, as well as for other magnetic characterization techniques, such as those based on X-rays or neutrons.

Upheaval of Negative Dielectric Permittivity and Semiconductor to Metallic Transition in a Thermally Induced Sn-Doped β-Ga2O3 (-201) Single Crystal.

Thermally induced dielectric and conductivity properties of an Sn-doped β-Ga2O3 (-201) single crystal were investigated by frequency-domain impedance spectroscopy in the frequency window from 100 Hz to 1 MHz with temperatures between 293 and 873 K. The (-201) plane-orientated single crystalline nature and the presence of an Sn dopant in β-Ga2O3 were confirmed by X-ray diffraction (XRD) and X-ray photoelectron (XPS) spectroscopy. Two different trends of impedance spectra have been discussed by the modulation of relaxation times and semiconductor to metallic transition after ∼723 K due to activation of a significant number of Sn dopants and their movements with temperature. The negative impedance values were encountered in the Nyquist plots (Z' vs Z″) after 573 K and constitute a reverse movement after 723 K with temperature. The average normalized change (ΔZ'/Δf)/Z0 of impedance exhibits a broad downward relaxation plateau near 723 K, indicating a weak electrical transition. The increases in the positive value of the dielectric constant (εr') below a percolating threshold temperature 573 K is attributed to the interfacial and dipolar polarizations, and the plasma oscillation of delocalized electrons governed by the Drude theory is responsible for the negative dielectric constant above 573 K. The 3D projections of the real dielectric constant create a sharp downward sinkhole near 723 K, indicating the existence of negative dielectric permittivity. The electrical conductivity dramatically changes its trends after 523 K and confirms a transition from hopping conduction (dielectric or semiconductor) following Jonscher's power law to metallic conduction by Drude theory. Below the percolating threshold temperature, a nonoverlapping small polaron tunneling conduction mechanism was unveiled with defect-induced activation energy of 0.21 eV. The Sn-doped β-Ga2O3 exhibits unique and tailored electromagnetic responses with temperatures that can be associated with a variety of applications in electromagnetic wave manipulations, cloaking devices, antennas, sensors, medical imaging, seismic wave propagation, etc.

Six-Degree-of-Freedom Pose Estimation Method for Multi-Source Feature Points Based on Fully Convolutional Neural Network

An object’s six-degree-of-freedom (6DoF) pose information has great importance in various fields. Existing methods of pose estimation usually detect two-dimensional (2D)-three-dimensional (3D) feature point pairs, and directly estimates the pose information through Perspective-n-Point (PnP) algorithms. However, this approach ignores the spatial association between pixels, making it difficult to obtain high-precision results. In order to apply pose estimation based on deep learning methods to real-world scenarios, we hope to design a method that is robust enough in more complex scenarios. Therefore, we introduce a method for 3D object pose estimation from color images based on farthest point sampling (FPS) and object 3D bounding box. This method detects the 2D projection of 3D feature points through a convolutional neural network, matches it with the 3D model of the object, and then uses the PnP algorithm to restore the feature point pair to the object pose. Due to the global nature of the bounding box, this approach can be considered effective even in partially occluded or complex environments. In addition, we propose a heatmap suppression method based on weighted coordinates to further improve the prediction accuracy of feature points and the accuracy of the PnP algorithm in solving the pose position. Compared with other algorithms, this method has higher accuracy and better robustness. Our method yielded 93.8% of the ADD(-s) metrics on the Linemod dataset and 47.7% of the ADD(-s) metrics on the Occlusion Linemod dataset. These results show that our method is more effective than existing methods in pose estimation of large objects.

Design of an experimental approach based on the contrast-to-noise ratio measurement for X-ray computed tomography parameters optimization applied to a carbon fiber-reinforced polymer materials scan

This paper proposes a systematic approach for the optimization of scan parameters for industrial X-ray computed tomography (XCT), as regard its specific application as diagnostic tool on carbon fiber-reinforced polymer materials (CFRP). This procedure allows the system operator to overcome suboptimal scan results due to a subjective choice of XCT parameters. In this work, XCT scan quality has been measured in terms of contrast-to-noise ratio (CNR) metric, by calculating it on collected XCT 2D projection images. A four-factor five-level central composite design (CCD) was implemented to perform experiments, and a quadratic polynomial model was chosen to describe the effects of XCT scanning parameters combination on CNR measurement and finally to predict optimal results. Analysis of variance was carried out to evaluate the significance of the model on the response, reporting a R2 of 97.1%, and response surface analyses were also performed for CNR optimization purposes. In order to validate the CCD results, different XCT parameters combinations, coming from the CCD analysis on projection images, were used to run different scans, and, as result, the optimal CNR predicted from the model was also reflected in an optimal CNR measured on the reconstructed XCT images.

Selective signal enhancement in Fourier space as a tool for discovering ultrastructural organization of macromolecules from in situ TEM

We present a Fourier transform (FT) based analytical method that allows to obtain of ultrastructural details from TEM images at sub-nanometer scale applying a selective filtering for singular macromolecule electron microscopy density information. It can be applied to high-pressure frozen, frozen hydrated and epoxy freeze substituted and embedded biological species. Both 2D projections and orthoslices from reconstructed tomograms can be used as a source of structural information. The key to the method is to select the macromolecule or organelle of interest with an accuracy of ≥ 7 – 3 nm (depending on pixel size of initial tilt series or singular image acquisition) and explore both the central low frequency FT intensity and diffraction regions to obtain the spatial structural organization and its dimensional characteristics, respectively. We also introduce a structure-specific selective mask FT filtering approach that can significantly improve image information even in poorly contrasted TEM of resin sections without heavy metal been used. The described method elucidates chromatin architecture without the need of averaging. A zigzag symmetry of 30 nm diameter chromatin fibers which in general is a controversial topic of research has been identified for C. elegans cells in vivo with sub-nanometer details being preserved in the images.

Predicting cell cycle stage from 3D single-cell nuclear-stained images.

The cell cycle governs the proliferation, differentiation, and regeneration of all eukaryotic cells. Profiling cell cycle dynamics is therefore central to basic and biomedical research spanning development, health, aging, and disease. However, current approaches to cell cycle profiling involve complex interventions that may confound experimental interpretation. To facilitate more efficient cell cycle annotation of microscopy data, we developed CellCycleNet, a machine learning (ML) workflow designed to simplify cell cycle staging with minimal experimenter intervention and cost. CellCycleNet accurately predicts cell cycle phase using only a fluorescent nuclear stain (DAPI) in fixed interphase cells. Using the Fucci2a cell cycle reporter system as ground truth, we collected two benchmarking image datasets and trained two ML models-a support vector machine (SVM) and a deep neural network-to classify nuclei as being in either the G1 or S/G2 phases of the cell cycle. Our results suggest that CellCycleNet outperforms state-of-the-art SVM models on each dataset individually. When trained on two image datasets simultaneously, CellCycleNet achieves the highest classification accuracy, with an improvement in AUROC of 0.08-0.09. The model also demonstrates excellent generalization across different microscopes, achieving an AUROC of 0.95. Overall, using features derived from 3D images, rather than 2D projections of those same images, significantly improves classification performance. We have released our image data, trained models, and software as a community resource.