Using In-frame Shear Constraints for Monocular Motion Segmentation of Rigid Bodies

Tourani, Siddharth; Krishna, K. Madhava

doi:10.1007/s10846-015-0195-1

Cited by 6 publications

(8 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 10 shows a benchmark comparison of the car sequences results using TbD-SfM (Table 9) and other state-of-the-art methods [38,39]. The results presented in the Table 10 shows that TbD-SfM achieves a lower segmentation error in scenes with two and three simultaneous motions in comparison to methods presented in [18,20,21,22,24,26,27,28,29,30,31,32]. TbD-SfM obtains a segmentation error of 0.07% for sequences involving three simultaneous motions.…”

Section: Resultsmentioning

confidence: 99%

“…Then, it is applied a clustering method based on the spectral clustering theory. Tourani et al [22] carried out the hypothesis generation using the RANSAC procedure. An over-segmentation is implemented by a long-term gestalt-inspired motion similarity constraints, into a multi-label Markov Random Field (MRF).…”

Section: Introductionmentioning

confidence: 99%

“…A comparison with other state-of-the-art techniques is provided in terms of segmentation and reprojection errors as proposed in [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]. The outliers ratio is also provided as an indicator of the number of segmented feature points.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Track-Before-Detect Framework-Based Vehicle Monocular Vision Sensors

Gonzalez

Rodríguez

Elouardi

2019

Sensors

View full text Add to dashboard Cite

This paper proposes a Track-before-Detect framework for a multibody motion segmentation (named TbD-SfM). Our contribution relies on a tightly coupled tracking before detection strategy intended to reduce the complexity of existing Multibody Structure from Motion approaches. Efforts were done towards an algorithm variant closer and aimed to a further embedded implementation for dynamic scene analysis while enhancing processing time performances. This generic motion segmentation approach can be transposed to several transportation sensor systems since no constraints are considered on segmented motions (6-DOF model). The tracking scheme is analyzed and its performance is evaluated under thorough experimental conditions including full-scale driving scenarios from known and available datasets. Results on challenging scenarios including the presence of multiple and simultaneous moving objects observed from a moving camera are reported and discussed.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Track-Before-Detect Framework-Based Vehicle Monocular Vision Sensors

Gonzalez

Rodríguez

Elouardi

2019

Sensors

View full text Add to dashboard Cite

show abstract

“…In the seminal work contributed by (Elhamifar and Vidal, 2009), trajectory points were modeled as sparse combination of evaluated trajectories. (Tourani and Krishna, 2016) used in frame shear constraints to generate and merge affine models, achieving state of art results in sparse motion segmentation using monocular camera. Recently many deep convolution nets have been used to learn motion labels (Rozantsev et al, 2014) (Fragkiadaki et al, 2015) (Tokmakov et al, 2016) for motion segmentation.…”

Section: Related Workmentioning

confidence: 99%

“…For outdoor robotic vision, the camera displacement is unavoidable. Although, this has been tackled in (Tourani and Krishna, 2016) where motion models are generated and merged using trajectory clustering into different motion affine subspaces. The moving object proposals generated from the prior model are sparse collection of points lying on the object, resulting into a sparse motion segmentation.…”

Section: Introductionmentioning

confidence: 99%

Joint Semantic and Motion Segmentation for Dynamic Scenes using Deep Convolutional Networks

Haque

Reddy

Krishna

2017

Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

View full text Add to dashboard Cite

Abstract:Dynamic scene understanding is a challenging problem and motion segmentation plays a crucial role in solving it. Incorporating semantics and motion enhances the overall perception of the dynamic scene. For applications of outdoor robotic navigation, joint learning methods have not been extensively used for extracting spatiotemporal features or adding different priors into the formulation. The task becomes even more challenging without stereo information being incorporated. This paper proposes an approach to fuse semantic features and motion clues using CNNs, to address the problem of monocular semantic motion segmentation. We deduce semantic and motion labels by integrating optical flow as a constraint with semantic features into dilated convolution network. The pipeline consists of three main stages i.e Feature extraction, Feature amplification and Multi Scale Context Aggregation to fuse the semantics and flow features. Our joint formulation shows significant improvements in monocular motion segmentation over the state of the art methods on challenging KITTI tracking dataset.

show abstract

MotionRFCN: Motion Segmentation Using Consecutive Dense Depth Maps

Liu

Wang

2019

PRICAI 2019: Trends in Artificial Intelligence

View full text Add to dashboard Cite

Using In-frame Shear Constraints for Monocular Motion Segmentation of Rigid Bodies

Cited by 6 publications

References 43 publications

Track-Before-Detect Framework-Based Vehicle Monocular Vision Sensors

Track-Before-Detect Framework-Based Vehicle Monocular Vision Sensors

Joint Semantic and Motion Segmentation for Dynamic Scenes using Deep Convolutional Networks

MotionRFCN: Motion Segmentation Using Consecutive Dense Depth Maps

Contact Info

Product

Resources

About