Skip to main content

Martin Cadik

Brno University of Technology, Faculty of Information Technology, Faculty Member

Followers

36

Following

17

Co-authors

6

Public Views

Graphics and vision researcher, associate professor http://cadik.posvete.cz/

less

University of Kansas

School of the Art Institute of Chicago

Duke University

Kansas State University

Fabrizio Galeazzi

Anglia Ruskin University

Miglena Stamberova

National Institute of Archaeology and Museum, Bulgarian Academy of Sciences

University of Melbourne

José d'Encarnação

Universidade de Coimbra

Manlio Vinciguerra

University College London

Fusun Balik Sanli

Yildiz Technical University

InterestsView All (10)

Uploads

Papers by Martin Cadik

Panorama Boundary Labeling with Temporal Coherence

Computer Graphics International, Jun 18, 2019

GeoPose3K: Mountain landscape dataset for camera pose estimation in outdoor environments

Image and Vision Computing, Oct 1, 2017

We introduce a new dataset called GeoPose3K 1 which contains over three thousand precise camera p... more We introduce a new dataset called GeoPose3K 1 which contains over three thousand precise camera poses of mountain landscape images. In addition to camera location and orientation, we provide data for the training and evaluation of computer vision methods and applications in the context of outdoor scenes; synthetic depth maps, normal maps, illumination simulation and semantic labels. In order to illustrate properties of the dataset, we compare results achieved by state-of-the-art visual geo-localization method on GeoPose3K with results achieved on an existing dataset for visual geo-localization. So as to foster research of computer vision algorithms for outdoor environments, several novel future use-cases of our new GeoPose3K dataset are proposed.

Vision UFormer: Long-range monocular absolute depth estimation

Computers & Graphics, Apr 1, 2023

Predicting photovoltaic power production using high-uncertainty weather forecasts

Applied Energy, Jun 1, 2023

Reinforced Labels: Multi-Agent Deep Reinforcement Learning for Point-Feature Label Placement

arXiv (Cornell University), Mar 2, 2023

Over the recent years, Reinforcement Learning combined with Deep Learning techniques has successf... more Over the recent years, Reinforcement Learning combined with Deep Learning techniques has successfully proven to solve complex problems in various domains, including robotics, self-driving cars, and finance. In this paper, we are introducing Reinforcement Learning (RL) to label placement, a complex task in data visualization that seeks optimal positioning for labels to avoid overlap and ensure legibility. Our novel point-feature label placement method utilizes Multi-Agent Deep Reinforcement Learning to learn the label placement strategy, the first machine-learning-driven labeling method, in contrast to the existing hand-crafted algorithms designed by human experts. To facilitate RL learning, we developed an environment where an agent acts as a proxy for a label, a short textual annotation that augments visualization. Our results show that the strategy trained by our method significantly outperforms the random strategy of an untrained agent and the compared methods designed by human experts in terms of completeness (i.e., the number of placed labels). The trade-off is increased computation time, making the proposed method slower than the compared methods. Nevertheless, our method is ideal for scenarios where the labeling can be computed in advance, and completeness is essential, such as cartographic maps, technical drawings, and medical atlases. Additionally, we conducted a user study to assess the perceived performance. The outcomes revealed that the participants considered the proposed method to be significantly better than the other examined methods. This indicates that the improved completeness is not just reflected in the quantitative metrics but also in the subjective evaluation by the participants.

Video Sequence Boundary Labeling with Temporal Coherence

Lecture Notes in Computer Science, 2019

We propose a method for video sequence boundary labeling which maintains the temporal coherence. ... more We propose a method for video sequence boundary labeling which maintains the temporal coherence. The method is based on two ideas. We limit the movement of the label boxes only to the horizontal direction, and reserve free space for the movement of the label boxes in the label layout. The proposed method is able to position label boxes in video sequence on a lower number of rows than existing methods, while at the same time, it minimizes the movement of label boxes. We conducted an extensive user experiment where the proposed method was ranked the best for panorama video sequences labeling compared to three existing methods.

PTRM: Perceived Terrain Realism Metrics

arXiv (Cornell University), Sep 10, 2019

Terrains are visually important and commonly used in computer graphics. While many algorithms for... more Terrains are visually important and commonly used in computer graphics. While many algorithms for their generation exist, it is difficult to assess the realism of a generated terrain. This paper presents a first step in the direction of perceptual evaluation of terrain models. We gathered and categorized several classes of real terrains and we generated synthetic terrains by using methods from computer graphics. We then conducted two large studies ranking the terrains perceptually and showing that the synthetic terrains are perceived as lacking realism as compared to the real ones. Then we provide insight into the features that affect the perceived realism by a quantitative evaluation based on localized geomorphology-based landform features (geomorphons) that categorize terrain structures such as valleys, ridges, hollows, etc. We show that the presence or absence of certain features have a significant perceptual effect. We then introduce Perceived Terrain Realism Metrics (PTRM); a perceptual metrics that estimates perceived realism of a terrain represented as a digital elevation map by relating distribution of terrain features with their perceived realism. We validated PTRM on real and synthetic data and compared it to the perceptual studies. To confirm the importance of the presence of these features, we used a generative deep neural network to transfer them between real terrains and synthetic ones and we performed another perceptual experiment that further confirmed their importance for perceived realism.

LandscapeAR: Large Scale Outdoor Augmented Reality by Matching Photographs with Terrain Models Using Learned Descriptors

Lecture Notes in Computer Science, 2020

Comparison of Semantic Segmentation Approaches for Horizon/Sky Line Detection

arXiv (Cornell University), May 21, 2018

Resource Efficient Mountainous Skyline Extraction using Shallow Learning

arXiv (Cornell University), Jul 22, 2021

Skyline plays a pivotal role in mountainous visual geo-localization and localization/navigation o... more Skyline plays a pivotal role in mountainous visual geo-localization and localization/navigation of planetary rovers/UAVs and virtual/augmented reality applications. We present a novel mountainous skyline detection approach where we adapt a shallow learning approach to learn a set of filters to discriminate between edges belonging to sky-mountain boundary and others coming from different regions. Unlike earlier approaches, which either rely on extraction of explicit feature descriptors and their classification, or fine-tuning general scene parsing deep networks for sky segmentation, our approach learns linear filters based on local structure analysis. At test time, for every candidate edge pixel, a single filter is chosen from the set of learned filters based on pixel's structure tensor, and then applied to the patch around it. We then employ dynamic programming to solve the shortest path problem for the resultant multistage graph to get the sky-mountain boundary. The proposed approach is computationally faster than earlier methods while providing comparable performance and is more suitable for resource constrained platforms e.g., mobile devices, planetary rovers and UAVs. We compare our proposed approach against earlier skyline detection methods using four different data sets. Our code is available at https://github.com/TouqeerAhmad/skyline detection.

Vision Uformer: Long-Range Monocular Absolute Depth Estimation

SSRN Electronic Journal

Pose Estimation from Line Correspondences using Direct Linear Transformation

arXiv (Cornell University), Aug 24, 2016

This work is concerned with camera pose estimation from correspondences of 3D/2D lines, i.e. with... more This work is concerned with camera pose estimation from correspondences of 3D/2D lines, i.e. with the Perspective-n-Line (PnL) problem. We focus on large line sets, which can be efficiently solved by methods using linear formulation of PnL. We propose a novel method 'DLT-Combined-Lines' based on the Direct Linear Transformation (DLT) algorithm, which benefits from a new combination of two existing DLT methods for pose estimation. The method represents 2D structure by lines, and 3D structure by both points and lines. The redundant 3D information reduces the minimum of required line correspondences to 5. A cornerstone of the method is a combined projection matrix estimated by the DLT algorithm. It contains multiple estimates of camera rotation and translation, which can be recovered after enforcing constraints of the matrix. Multiplicity of the estimates is exploited to improve accuracy of the proposed method. For large line sets (10 and more), the method achieves state-of-the-art accuracy on synthetic data even under strong image noise. Moreover, it is the most accurate method on real world data, outperforming the state-of-the-art by a large margin. The proposed method is also highly computationally effective, estimating the pose of 1000 lines in 12 ms on a desktop computer.

Panorama Boundary Labeling with Temporal Coherence

PTRM: Perceived Terrain Realism Metric

ACM Transactions on Applied Perception, 2022

Terrains are visually prominent and commonly needed objects in many computer graphics application... more Terrains are visually prominent and commonly needed objects in many computer graphics applications. While there are many algorithms for synthetic terrain generation, it is rather difficult to assess the realism of a generated output. This article presents a first step toward the direction of perceptual evaluation for terrain models. We gathered and categorized several classes of real terrains, and we generated synthetic terrain models using computer graphics methods. The terrain geometries were rendered by using the same texturing, lighting, and camera position. Two studies on these image sets were conducted, ranking the terrains perceptually, and showing that the synthetic terrains are perceived as lacking realism compared to the real ones. We provide insight into the features that affect the perceived realism by a quantitative evaluation based on localized geomorphology-based landform features (geomorphons) that categorize terrain structures such as valleys, ridges, hollows, and s...

ACM Transactions on Graphics, 2021

Many algorithms for virtual tree generation exist, but the visual realism of the 3D models is unk... more Many algorithms for virtual tree generation exist, but the visual realism of the 3D models is unknown. This problem is usually addressed by performing limited user studies or by a side-by-side visual comparison. We introduce an automated system for realism assessment of the tree model based on their perception. We conducted a user study in which 4,000 participants compared over one million pairs of images to collect subjective perceptual scores of a large dataset of virtual trees. The scores were used to train two neural-network-based predictors. A view independent ICTreeF uses the tree model's geometric features that are easy to extract from any model. The second is ICTreeI that estimates the perceived visual realism of a tree from its image. Moreover, to provide an insight into the problem, we deduce intrinsic attributes and evaluate which features make trees look like real trees. In particular, we show that branching angles, length of branches, and widths are critical for per...

HDR Video Metrics

High Dynamic Range Video, 2017

In this chapter we present an introduction to HDR image and video quality assessment fields. We d... more In this chapter we present an introduction to HDR image and video quality assessment fields. We discuss full-, no-, and reduced-reference metrics, including perceptually motivated methods. We describe two existing full-reference HDR video quality metrics in detail. Furthermore, we introduce the emerging field of data-driven metrics. Finally, we conclude with the outlook of future development and research.

ICTree: Automatic Perceptual Metrics for Tree Models

Fig. 1. A tree’s perceived visual realism guides the user during a real-time interaction with a 3... more Fig. 1. A tree’s perceived visual realism guides the user during a real-time interaction with a 3D tree modeling system. Two metrics are provided, a viewindependent ICTreeF uses the 3D geometry solely (light green bar, left), and ICTreeI uses a 2D image of the tree (dark green bar, right). The initial tree (left) is estimated as not visually pleasing, as indicated by the low values of the ICTreeI and ICTreeF. After adding branches and randomizing their values, the perceived realism has improved significantly (middle). The user then decreases the branching angles’ randomization and the branch lengths that impact the perceived realism (right).

Automated outdoor depth-map generation and alignment

Computers & Graphics, 2018

Image enhancement tasks can highly benefit from depth information, but the direct estimation of o... more Image enhancement tasks can highly benefit from depth information, but the direct estimation of outdoor depth maps is difficult due to vast object distances. This paper presents a fully automatic framework for model-based generation of outdoor depth maps and its applications to image enhancements. We leverage 3D terrain models and camera pose estimation techniques to render approximate depth maps without resorting to manual alignment. Potential local misalignments, resulting from insufficient model details and rough registrations, are eliminated with our novel free-form warping. We first align synthetic depth edges with photo edges using the as-rigid-as-possible image registration and further refine the shape of the edges using the tight trimap-based alpha matting. The resulting synthetic depth maps are accurate, calibrated in the absolute distance. We demonstrate their benefit in image enhancement techniques including reblurring, depth-of-field simulation, haze removal, and guided texture synthesis.

Temporally stable boundary labeling for interactive and non-interactive dynamic scenes

Computers & Graphics, 2020

We propose two novel temporally stable screen-space labeling methods for dynamic scenes. The firs... more We propose two novel temporally stable screen-space labeling methods for dynamic scenes. The first one is suitable for offline processing of the entire interaction or the video in advance. The second method is designed for interactive applications. The main idea of our proposed methods is to minimize the vertical and horizontal movement of the labels during the interaction with the scene (e.g., zooming or translating the camera). According to the results of quantitative evaluation, our labeling is more stable during the interaction than labeling produced by the current state of the art. Moreover, participants of a comprehensive user study declared that the labeling produced by the proposed methods allows them to follow moving labels significantly more accurately, and it is significantly more pleasing than with previously published methods. Furthermore, the proposed methods can be extended by the prominence of the features and easily parameterized to fit different requirements to the label layout.

State-of-the-art in visual geo-localization

Pattern Analysis and Applications, 2017

Large-scale visual geo-localization has recently gained a lot of attention in computer vision res... more Large-scale visual geo-localization has recently gained a lot of attention in computer vision research and new methods are proposed steadily. However, surveys of visual geo-localization methods are rare and they focus mainly on city-scale localization methods. We present a comprehensive and balanced study of existing visual geo-localization domains, including city-scale, global approaches and methods for natural environments. We overview the methods to show their pros and cons, application domains, datasets, as well as evaluation techniques. We categorize the reviewed methods by two criteria. The first is the type of data the method uses for geo-location estimation. The second criterion is the target environment for which the method has been proposed and validated. Based on this categorization we analyze important conditions that must be considered while solving geo-localization problems. Each category is in a different state of research-while city-scale image-based methods received a lot of attention, other categories like natural environments using cross-domain data sources are still challenging problems under active research. Future research of large-scale visual geo-localization is discussed, primarily the challenging and new research category-geo-localization in natural environments.

Panorama Boundary Labeling with Temporal Coherence

Computer Graphics International, Jun 18, 2019

GeoPose3K: Mountain landscape dataset for camera pose estimation in outdoor environments

Image and Vision Computing, Oct 1, 2017

We introduce a new dataset called GeoPose3K 1 which contains over three thousand precise camera p... more We introduce a new dataset called GeoPose3K 1 which contains over three thousand precise camera poses of mountain landscape images. In addition to camera location and orientation, we provide data for the training and evaluation of computer vision methods and applications in the context of outdoor scenes; synthetic depth maps, normal maps, illumination simulation and semantic labels. In order to illustrate properties of the dataset, we compare results achieved by state-of-the-art visual geo-localization method on GeoPose3K with results achieved on an existing dataset for visual geo-localization. So as to foster research of computer vision algorithms for outdoor environments, several novel future use-cases of our new GeoPose3K dataset are proposed.

Vision UFormer: Long-range monocular absolute depth estimation

Computers & Graphics, Apr 1, 2023

Predicting photovoltaic power production using high-uncertainty weather forecasts

Applied Energy, Jun 1, 2023

Reinforced Labels: Multi-Agent Deep Reinforcement Learning for Point-Feature Label Placement

arXiv (Cornell University), Mar 2, 2023

Over the recent years, Reinforcement Learning combined with Deep Learning techniques has successf... more Over the recent years, Reinforcement Learning combined with Deep Learning techniques has successfully proven to solve complex problems in various domains, including robotics, self-driving cars, and finance. In this paper, we are introducing Reinforcement Learning (RL) to label placement, a complex task in data visualization that seeks optimal positioning for labels to avoid overlap and ensure legibility. Our novel point-feature label placement method utilizes Multi-Agent Deep Reinforcement Learning to learn the label placement strategy, the first machine-learning-driven labeling method, in contrast to the existing hand-crafted algorithms designed by human experts. To facilitate RL learning, we developed an environment where an agent acts as a proxy for a label, a short textual annotation that augments visualization. Our results show that the strategy trained by our method significantly outperforms the random strategy of an untrained agent and the compared methods designed by human experts in terms of completeness (i.e., the number of placed labels). The trade-off is increased computation time, making the proposed method slower than the compared methods. Nevertheless, our method is ideal for scenarios where the labeling can be computed in advance, and completeness is essential, such as cartographic maps, technical drawings, and medical atlases. Additionally, we conducted a user study to assess the perceived performance. The outcomes revealed that the participants considered the proposed method to be significantly better than the other examined methods. This indicates that the improved completeness is not just reflected in the quantitative metrics but also in the subjective evaluation by the participants.

Video Sequence Boundary Labeling with Temporal Coherence

Lecture Notes in Computer Science, 2019

We propose a method for video sequence boundary labeling which maintains the temporal coherence. ... more We propose a method for video sequence boundary labeling which maintains the temporal coherence. The method is based on two ideas. We limit the movement of the label boxes only to the horizontal direction, and reserve free space for the movement of the label boxes in the label layout. The proposed method is able to position label boxes in video sequence on a lower number of rows than existing methods, while at the same time, it minimizes the movement of label boxes. We conducted an extensive user experiment where the proposed method was ranked the best for panorama video sequences labeling compared to three existing methods.

PTRM: Perceived Terrain Realism Metrics

arXiv (Cornell University), Sep 10, 2019

Terrains are visually important and commonly used in computer graphics. While many algorithms for... more Terrains are visually important and commonly used in computer graphics. While many algorithms for their generation exist, it is difficult to assess the realism of a generated terrain. This paper presents a first step in the direction of perceptual evaluation of terrain models. We gathered and categorized several classes of real terrains and we generated synthetic terrains by using methods from computer graphics. We then conducted two large studies ranking the terrains perceptually and showing that the synthetic terrains are perceived as lacking realism as compared to the real ones. Then we provide insight into the features that affect the perceived realism by a quantitative evaluation based on localized geomorphology-based landform features (geomorphons) that categorize terrain structures such as valleys, ridges, hollows, etc. We show that the presence or absence of certain features have a significant perceptual effect. We then introduce Perceived Terrain Realism Metrics (PTRM); a perceptual metrics that estimates perceived realism of a terrain represented as a digital elevation map by relating distribution of terrain features with their perceived realism. We validated PTRM on real and synthetic data and compared it to the perceptual studies. To confirm the importance of the presence of these features, we used a generative deep neural network to transfer them between real terrains and synthetic ones and we performed another perceptual experiment that further confirmed their importance for perceived realism.

LandscapeAR: Large Scale Outdoor Augmented Reality by Matching Photographs with Terrain Models Using Learned Descriptors

Lecture Notes in Computer Science, 2020

Comparison of Semantic Segmentation Approaches for Horizon/Sky Line Detection

arXiv (Cornell University), May 21, 2018

Resource Efficient Mountainous Skyline Extraction using Shallow Learning

arXiv (Cornell University), Jul 22, 2021

Skyline plays a pivotal role in mountainous visual geo-localization and localization/navigation o... more Skyline plays a pivotal role in mountainous visual geo-localization and localization/navigation of planetary rovers/UAVs and virtual/augmented reality applications. We present a novel mountainous skyline detection approach where we adapt a shallow learning approach to learn a set of filters to discriminate between edges belonging to sky-mountain boundary and others coming from different regions. Unlike earlier approaches, which either rely on extraction of explicit feature descriptors and their classification, or fine-tuning general scene parsing deep networks for sky segmentation, our approach learns linear filters based on local structure analysis. At test time, for every candidate edge pixel, a single filter is chosen from the set of learned filters based on pixel's structure tensor, and then applied to the patch around it. We then employ dynamic programming to solve the shortest path problem for the resultant multistage graph to get the sky-mountain boundary. The proposed approach is computationally faster than earlier methods while providing comparable performance and is more suitable for resource constrained platforms e.g., mobile devices, planetary rovers and UAVs. We compare our proposed approach against earlier skyline detection methods using four different data sets. Our code is available at https://github.com/TouqeerAhmad/skyline detection.

Vision Uformer: Long-Range Monocular Absolute Depth Estimation

SSRN Electronic Journal

Pose Estimation from Line Correspondences using Direct Linear Transformation

arXiv (Cornell University), Aug 24, 2016

This work is concerned with camera pose estimation from correspondences of 3D/2D lines, i.e. with... more This work is concerned with camera pose estimation from correspondences of 3D/2D lines, i.e. with the Perspective-n-Line (PnL) problem. We focus on large line sets, which can be efficiently solved by methods using linear formulation of PnL. We propose a novel method 'DLT-Combined-Lines' based on the Direct Linear Transformation (DLT) algorithm, which benefits from a new combination of two existing DLT methods for pose estimation. The method represents 2D structure by lines, and 3D structure by both points and lines. The redundant 3D information reduces the minimum of required line correspondences to 5. A cornerstone of the method is a combined projection matrix estimated by the DLT algorithm. It contains multiple estimates of camera rotation and translation, which can be recovered after enforcing constraints of the matrix. Multiplicity of the estimates is exploited to improve accuracy of the proposed method. For large line sets (10 and more), the method achieves state-of-the-art accuracy on synthetic data even under strong image noise. Moreover, it is the most accurate method on real world data, outperforming the state-of-the-art by a large margin. The proposed method is also highly computationally effective, estimating the pose of 1000 lines in 12 ms on a desktop computer.

Panorama Boundary Labeling with Temporal Coherence

PTRM: Perceived Terrain Realism Metric

ACM Transactions on Applied Perception, 2022

Terrains are visually prominent and commonly needed objects in many computer graphics application... more Terrains are visually prominent and commonly needed objects in many computer graphics applications. While there are many algorithms for synthetic terrain generation, it is rather difficult to assess the realism of a generated output. This article presents a first step toward the direction of perceptual evaluation for terrain models. We gathered and categorized several classes of real terrains, and we generated synthetic terrain models using computer graphics methods. The terrain geometries were rendered by using the same texturing, lighting, and camera position. Two studies on these image sets were conducted, ranking the terrains perceptually, and showing that the synthetic terrains are perceived as lacking realism compared to the real ones. We provide insight into the features that affect the perceived realism by a quantitative evaluation based on localized geomorphology-based landform features (geomorphons) that categorize terrain structures such as valleys, ridges, hollows, and s...

ACM Transactions on Graphics, 2021

Many algorithms for virtual tree generation exist, but the visual realism of the 3D models is unk... more Many algorithms for virtual tree generation exist, but the visual realism of the 3D models is unknown. This problem is usually addressed by performing limited user studies or by a side-by-side visual comparison. We introduce an automated system for realism assessment of the tree model based on their perception. We conducted a user study in which 4,000 participants compared over one million pairs of images to collect subjective perceptual scores of a large dataset of virtual trees. The scores were used to train two neural-network-based predictors. A view independent ICTreeF uses the tree model's geometric features that are easy to extract from any model. The second is ICTreeI that estimates the perceived visual realism of a tree from its image. Moreover, to provide an insight into the problem, we deduce intrinsic attributes and evaluate which features make trees look like real trees. In particular, we show that branching angles, length of branches, and widths are critical for per...

HDR Video Metrics

High Dynamic Range Video, 2017

In this chapter we present an introduction to HDR image and video quality assessment fields. We d... more In this chapter we present an introduction to HDR image and video quality assessment fields. We discuss full-, no-, and reduced-reference metrics, including perceptually motivated methods. We describe two existing full-reference HDR video quality metrics in detail. Furthermore, we introduce the emerging field of data-driven metrics. Finally, we conclude with the outlook of future development and research.

ICTree: Automatic Perceptual Metrics for Tree Models

Fig. 1. A tree’s perceived visual realism guides the user during a real-time interaction with a 3... more Fig. 1. A tree’s perceived visual realism guides the user during a real-time interaction with a 3D tree modeling system. Two metrics are provided, a viewindependent ICTreeF uses the 3D geometry solely (light green bar, left), and ICTreeI uses a 2D image of the tree (dark green bar, right). The initial tree (left) is estimated as not visually pleasing, as indicated by the low values of the ICTreeI and ICTreeF. After adding branches and randomizing their values, the perceived realism has improved significantly (middle). The user then decreases the branching angles’ randomization and the branch lengths that impact the perceived realism (right).

Automated outdoor depth-map generation and alignment

Computers & Graphics, 2018

Image enhancement tasks can highly benefit from depth information, but the direct estimation of o... more Image enhancement tasks can highly benefit from depth information, but the direct estimation of outdoor depth maps is difficult due to vast object distances. This paper presents a fully automatic framework for model-based generation of outdoor depth maps and its applications to image enhancements. We leverage 3D terrain models and camera pose estimation techniques to render approximate depth maps without resorting to manual alignment. Potential local misalignments, resulting from insufficient model details and rough registrations, are eliminated with our novel free-form warping. We first align synthetic depth edges with photo edges using the as-rigid-as-possible image registration and further refine the shape of the edges using the tight trimap-based alpha matting. The resulting synthetic depth maps are accurate, calibrated in the absolute distance. We demonstrate their benefit in image enhancement techniques including reblurring, depth-of-field simulation, haze removal, and guided texture synthesis.

Temporally stable boundary labeling for interactive and non-interactive dynamic scenes

Computers & Graphics, 2020

We propose two novel temporally stable screen-space labeling methods for dynamic scenes. The firs... more We propose two novel temporally stable screen-space labeling methods for dynamic scenes. The first one is suitable for offline processing of the entire interaction or the video in advance. The second method is designed for interactive applications. The main idea of our proposed methods is to minimize the vertical and horizontal movement of the labels during the interaction with the scene (e.g., zooming or translating the camera). According to the results of quantitative evaluation, our labeling is more stable during the interaction than labeling produced by the current state of the art. Moreover, participants of a comprehensive user study declared that the labeling produced by the proposed methods allows them to follow moving labels significantly more accurately, and it is significantly more pleasing than with previously published methods. Furthermore, the proposed methods can be extended by the prominence of the features and easily parameterized to fit different requirements to the label layout.

State-of-the-art in visual geo-localization

Pattern Analysis and Applications, 2017

Large-scale visual geo-localization has recently gained a lot of attention in computer vision res... more Large-scale visual geo-localization has recently gained a lot of attention in computer vision research and new methods are proposed steadily. However, surveys of visual geo-localization methods are rare and they focus mainly on city-scale localization methods. We present a comprehensive and balanced study of existing visual geo-localization domains, including city-scale, global approaches and methods for natural environments. We overview the methods to show their pros and cons, application domains, datasets, as well as evaluation techniques. We categorize the reviewed methods by two criteria. The first is the type of data the method uses for geo-location estimation. The second criterion is the target environment for which the method has been proposed and validated. Based on this categorization we analyze important conditions that must be considered while solving geo-localization problems. Each category is in a different state of research-while city-scale image-based methods received a lot of attention, other categories like natural environments using cross-domain data sources are still challenging problems under active research. Future research of large-scale visual geo-localization is discussed, primarily the challenging and new research category-geo-localization in natural environments.