Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,291 papers · 148 categories

Trend · papers per month

316394125 · May 202619922001200920182026
48 results for Optical flow

Novel video prediction method for complex urban scenes using optical flow.

problem Making accurate future frame predictions in complex urban scenes.
method Optical flow conditioned method using video sequences and optical flow sequences.
result Empirical evaluations show the effectiveness of the method on KITTI and Cityscapes datasets.

The paper solves the isoperimetric problem in Riemannian optical geometry, proving circles minimize lengths with area constraints.

problem Optical geometry of static spherically symmetric spacetimes.
method Applying isoperimetric problem results to curves in Riemannian optical geometry.
result Length-minimizing curves with area constraints are circles, with implications for photon spheres.

In this article we prove two formulas for the topological entropy of an F-optical Hamiltonian flow induced by a C^{\infty} Hamiltonian, where F is a Lagrangian distribution. In these formulas, we calculate the topological entropy as the exponential growth rate of the average of the determinant of the differential of th…

2000-10-05abs ↗pdf ↗

This paper explores synthetic data for training networks in stereo and optical flow tasks.

problem Creating accurate training data for stereo and optical flow tasks is challenging.
method Promotes the use of synthetically generated data and evaluates its impact on network performance.
result Synthetically generated data improves network performance and generalization.

COMBO network improves optical flow estimation by combining deep learning with brightness constancy.

problem Optical flow estimation using deep learning requires complex training schemes.
method COMBO network explicitly exploits brightness constancy and combines it with a data-driven approach.
result COMBO network outperforms state-of-the-art methods on various benchmarks.

Two machine learning applications for IP/Optical networks: traffic prediction and optical path performance.

problem Agile resource management and optical path performance prediction in IP/Optical networks.
method Machine learning for traffic prediction and optical performance prediction using SDN controllers.
result Efficient implementation of SDN controllers for agile resource management and optical path performance prediction.

New model predicts radiative properties of nanoparticle layers with high accuracy and uncertainty.

problem Predicting radiative properties of nanoparticle embedded layers accurately and with uncertainty.
method Conditional normalizing flows learn conditional distributions of optical outputs given input parameters.
result The model achieves high predictive accuracy and reliable uncertainty estimates.

New approach uses optical flow error for lightning prediction.

problem Improving lightning prediction accuracy using machine learning.
method Optical flow error as feature, manual convolution for spatial proximity, various features considered.
result High accuracy of 96% for 15-minute predictions, decreasing to 83% for 5-hour predictions.

Improved object segmentation and tracking in video using optical flow and initial state conditioning.

problem Challenges in fully unsupervised object-centric learning from video data.
method Weakly-supervised approach using optical flow and initial state conditioning.
result Conditioning the model on simple object location cues significantly improves instance segmentation in realistic synthetic data.

This paper uses deep learning to estimate flow fields from OCT images for laser ablation control.

problem Automatic control of laser bone ablation using 4D OCT images.
method Semi-supervised convolutional neural network for 2.5D scene flow estimation.
result Scene flow estimation enables markerless tracking and automated laser ablation control.

We consider an optical hypersurface ΣΣ in the cotangent bundle τ:TMMτ:T^*M\to M of a closed manifold MM endowed with a twisted symplectic structure. We show that if the characteristic foliation of ΣΣ is Anosov, then a smooth 1-form θθ on MM is exact if and only τθτ^*θ has zero integral over every closed characteristi…

2005-08-17abs ↗pdf ↗

Combining visual attention with deep reinforcement learning improves sample efficiency.

problem Improving sample efficiency in deep reinforcement learning.
method Visual selective attention mechanism implemented using optical flow and batch normalization.
result Visual selective attention leads to improvements in sample efficiency on Atari games.

SVM with local features improves human action recognition.

problem Improving human action recognition in videos.
method Local appearance and motion features extracted using CNNs, concatenated, and used with SVM for classification.
result SVM with local features outperforms previous methods on benchmark datasets.

We consider a triality between the Zermelo navigation problem, the geodesic flow on a Finslerian geometry of Randers type, and spacetimes in one dimension higher admitting a timelike conformal Killing vector field. From the latter viewpoint, the data of the Zermelo problem are encoded in a (conformally) Painleve-Gullst…

2008-11-18abs ↗pdf ↗

This work improves optic disc and cup segmentation for glaucoma detection.

problem Automatic segmentation of optic disc and cup on eye fundus images for glaucoma diagnosis.
method Modification of U-Net convolutional neural network.
result Our method achieves comparable quality to state-of-the-art methods, with faster prediction times.

Paper tackles depth estimation and optic disc-cup segmentation from color fundus images.

problem Depth estimation and optic disc-cup segmentation from color fundus images.
method Uses fully convolutional networks for monocular retinal depth estimation and optic disc-cup segmentation.
result Demonstrates improved accuracy in depth estimation and optic disc-cup segmentation.

Deep learning enhances optical microscopy and image reconstruction.

problem Improving image data transformations in optical microscopy.
method Application of deep learning methods on optical microscopy and image reconstruction.
result Deep learning enables new transformations among different modes and modalities of microscopic imaging.

GANPOP uses deep learning to estimate optical properties from single images, improving accuracy over existing methods.

problem Estimating optical properties from single wide-field images.
method Conditional generative adversarial networks trained on paired images and optical property maps.
result GANPOP estimates optical properties with 58% higher accuracy than single-snapshot optical property technique in human gastrointestinal specimens.

Paper develops a deforestation detection system using optical and SAR data.

problem Detecting tree-loss in dense forests using satellite data.
method Combines optical and SAR data, uses KL expansion for anomaly detection, and Hidden Markov Model for classification.
result Hybrid method achieves high accuracy and robustness in sparse optical data.

Study of optical geometries with intrinsic torsion in general relativity.

problem Understanding null line distributions and their properties in Lorentzian manifolds.
method Investigation of intrinsic torsion and congruences of null curves, extending to generalized optical geometries.
result Characterization of conformal properties of null line distributions and congruences.

New method speeds up galaxy analysis from hours to seconds.

problem Infeasibility of state-of-the-art SED analyses for large surveys.
method Amortized Neural Posterior Estimation (ANPE) for scalable Bayesian inference.
result Posterior distributions of 12 model parameters estimated in seconds per galaxy.

Paper develops machine learning to translate SAR to optical images for easier interpretation.

problem Difficulty in human interpretation of SAR images due to non-adapted human vision to microwave scattering.
method Develops a novel reciprocal GAN scheme to train machine intelligence on co-registered SAR and optical images.
result The proposed translation network works well under various SAR and optical image resolutions and polarizations.

Paper introduces TAP-Vid, a benchmark for tracking any point in videos.

problem Tackles the problem of tracking arbitrary physical points on surfaces over longer video clips.
method Formalizes the problem as TAP, introduces TAP-Vid benchmark, uses crowdsourced pipeline with optical flow estimates, proposes TAP-Net model.
result TAP-Net outperforms all prior methods on TAP-Vid benchmark when trained on synthetic data.

Express Wavenet reduces neural network parameters to 1% of standard networks.

problem Optical neural networks with high parameter count.
method Wavelet modulation, random shift wavelets, expressway structure.
result Express Wavenet achieves high accuracy with significantly fewer parameters.

Designs chiral photonic structures using machine learning for efficient optical properties.

problem Optimizing chiral photonic nanostructures for light-matter interactions.
method Evolutionary algorithm and neural network approach for rapid optimization.
result Frequency-dependent modification in reflected light's degree of circular polarization.

New methods for constructing null fluid metrics and solving optical lift conjectures.

problem Constructing null fluid metrics and solving optical lift conjectures.
method Explicit parameterization of null fluid metrics under Kerr type optical structures.
result New explicit metrics, including Kerr black holes and Ricci flat examples.

Hamilton proved a theorem about focusing light rays, leading to new mathematical concepts.

problem Focusing light rays to a single point using optical instruments.
method Original proof and symplectic geometry proof of the Malus-Dupin theorem.
result A family of light rays can be focused to a single point if they are rectangular before entering the instrument.