Learning from scratch a confidence measure

Poggi, Matteo; Mattoccia, Stefano

doi:10.5244/c.30.46

Cited by 89 publications

(126 citation statements)

References 0 publications

Supporting

Mentioning

123

Contrasting

Unclassified

Order By: Relevance

“…The related disparity maps are generated from the cost volumes via Winner Take All (WTA) strategy. The results are compared against the state-of-the-art confidence estimation methods CCNN [18], LFN [20] and LGC-Net [21]. To allow a fair comparison, all examined methods have been trained on the same data, following the procedure described in Section 3.3.…”

Section: Resultsmentioning

confidence: 99%

“…Finally, the approaches within the third group map the whole confidence estimation process to convolutional neural networks. For this purpose, [1,18,19] utilise square patches extracted from disparity maps and centred on a pixel of interest to determine its confidence. [14] in addition, proposes to stack two of those patches, one from the left, one from the right image, in order to introduce the idea of left-right-consistency.…”

Section: Confidence Estimationmentioning

confidence: 99%

“…The third and last part of the network consists of two fully-connected layers: One with 16 nodes followed by a ReLU activation and one with a single node and a subsequent sigmoid activation function, which provides the final confidence estimation. Following a common procedure [18,32], the fully-connected layers are replaced by convolutional layers, transforming the proposed architecture to a fully Figure 4: Visual proof of the necessity for a refined evaluation method. Since the original curve estimation does not sample within ranges of the same confidence score, it may suffer from a high discretisation error (here especially present in the upper right part).…”

Section: Architecturementioning

confidence: 99%

See 2 more Smart Citations

CNN-Based Cost Volume Analysis as Confidence Measure for Dense Matching

Mehltretter

Heipke

2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

View full text Add to dashboard Cite

Due to its capability to identify erroneous disparity assignments in dense stereo matching, confidence estimation is beneficial for a wide range of applications, e.g. autonomous driving, which needs a certain degree of confidence as mandatory prerequisite. Especially, the introduction of deep learning based methods resulted in an increasing popularity of this field in recent years, caused by a significantly improved accuracy. Despite this remarkable development, most of these methods rely on features learned from disparity maps only, not taking into account the corresponding 3-dimensional cost volumes. However, it was already demonstrated that with conventional methods based on hand-crafted features this additional information can be used to further increase the accuracy. In order to combine the advantages of deep learning and cost volume based features, in this paper, we propose a novel Convolutional Neural Network (CNN) architecture to directly learn features for confidence estimation from volumetric 3D data. An extensive evaluation on three datasets using three common dense stereo matching techniques demonstrates the generality and state-of-the-art accuracy of the proposed method.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Confidence Estimationmentioning

confidence: 99%

Section: Architecturementioning

confidence: 99%

See 1 more Smart Citation

CNN-Based Cost Volume Analysis as Confidence Measure for Dense Matching

Mehltretter

Heipke

2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

View full text Add to dashboard Cite

show abstract

“…A wide range of 3D reconstruction approaches estimate confidence values for depth hypotheses which are then later used for adaptive fusion. All these approaches typically use either handcrafted confidence weights [18,48,30] rather than learning them intrinsically from data or they learn only 2D score map without learning their 3D fusion [37,45,44,50]. Semantic 3D Reconstruction and Scene Completion.…”

Section: Related Workmentioning

confidence: 99%

Learned Semantic Multi-Sensor Depth Map Fusion

Rozumnyi¹,

Cherabier²,

Pollefeys³

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

View full text Add to dashboard Cite

Volumetric depth map fusion based on truncated signed distance functions has become a standard method and is used in many 3D reconstruction pipelines. In this paper, we are generalizing this classic method in multiple ways: 1) Semantics: Semantic information enriches the scene representation and is incorporated into the fusion process. 2) Multi-Sensor: Depth information can originate from different sensors or algorithms with very different noise and outlier statistics which are considered during data fusion. 3) Scene denoising and completion: Sensors can fail to recover depth for certain materials and light conditions, or data is missing due to occlusions. Our method denoises the geometry, closes holes and computes a watertight surface for every semantic class. 4) Learning: We propose a neural network reconstruction method that unifies all these properties within a single powerful framework. Our method learns sensor or algorithm properties jointly with semantic depth fusion and scene completion and can also be used as an expert system, e.g. to unify the strengths of various photometric stereo algorithms. Our approach is the first to unify all these properties. Experimental evaluations on both synthetic and real data sets demonstrate clear improvements.

show abstract

“…To overcome the requirement for a sufficient amount of training data, the recent trend is to use only weak supervision. Tonioni et al [9] generate their training data from a traditional formulation [10], but estimate a confidence score for the established matches with another CNN [11]. For training their regression network, the loss function combines the confidence weight to penalize deviations to their generated training data with an additional smoothness constraint on the solution.…”

Section: Related Workmentioning

confidence: 99%

Self-Supervised Learning for Stereo Reconstruction on Aerial Images

Knöbelreiter

Vogel

Pock

2018

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

Recent developments established deep learning as an inevitable tool to boost the performance of dense matching and stereo estimation. On the downside, learning these networks requires a substantial amount of training data to be successful. Consequently, the application of these models outside of the laboratory is far from straight forward. In this work we propose a self-supervised training procedure that allows us to adapt our network to the specific (imaging) characteristics of the dataset at hand, without the requirement of external ground truth data. We instead generate interim training data by running our intermediate network on the whole dataset, followed by conservative outlier filtering. Bootstrapped from a pre-trained version of our hybrid CNN-CRF model, we alternate the generation of training data and network training. With this simple concept we are able to lift the completeness and accuracy of the pre-trained version significantly. We also show that our final model compares favorably to other popular stereo estimation algorithms on an aerial dataset.Index Termslarge scale 3D, dense matching, CNN

show abstract

Learning from scratch a confidence measure

Cited by 89 publications

References 0 publications

CNN-Based Cost Volume Analysis as Confidence Measure for Dense Matching

CNN-Based Cost Volume Analysis as Confidence Measure for Dense Matching

Learned Semantic Multi-Sensor Depth Map Fusion

Self-Supervised Learning for Stereo Reconstruction on Aerial Images

Contact Info

Product

Resources

About