Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userZoom and Learn: Generalizing Deep Stereo Matching to Novel Domains
59
0
0.0
(
0
)
Added by
Jiahao Pang
Publication date
2018
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Despite the recent success of stereo matching with convolutional neural networks (CNNs), it remains arduous to generalize a pre-trained deep stereo model to a novel domain. A major difficulty is to collect accurate ground-truth disparities for stereo pairs in the target domain. In this work, we propose a self-adaptation approach for CNN training, utilizing both synthetic training data (with ground-truth disparities) and stereo pairs in the new domain (without ground-truths). Our method is driven by two empirical observations. By feeding real stereo pairs of different domains to stereo models pre-trained with synthetic data, we see that: i) a pre-trained model does not generalize well to the new domain, producing artifacts at boundaries and ill-posed regions; however, ii) feeding an up-sampled stereo pair leads to a disparity map with extra details. To avoid i) while exploiting ii), we formulate an iterative optimization problem with graph Laplacian regularization. At each iteration, the CNN adapts itself better to the new domain: we let the CNN learn its own higher-resolution output; at the meanwhile, a graph Laplacian regularization is imposed to discriminatively keep the desired edges while smoothing out the artifacts. We demonstrate the effectiveness of our method in two domains: daily scenes collected by smartphone cameras, and street views captured in a driving car.
rate research
Read More
This paper shows that when applying machine learning to digital zoom for photography, it is beneficial to use real, RAW sensor data for training. Existing learning-based super-resolution methods do not use real sensor data, instead operating on RGB images. In practice, these approaches result in loss of detail and accuracy in their digitally zoomed output when zooming in on distant image regions. We also show that synthesizing sensor data by resampling high-resolution RGB images is an oversimplified approximation of real sensor data and noise, resulting in worse image quality. The key barrier to using real sensor data for training is that ground truth high-resolution imagery is missing. We show how to obtain the ground-truth data with optically zoomed images and contribute a dataset, SR-RAW, for real-world computational zoom. We use SR-RAW to train a deep network with a novel contextual bilateral loss (CoBi) that delivers critical robustness to mild misalignment in input-output image pairs. The trained network achieves state-of-the-art performance in 4X and 8X computational zoom.
To reduce the human efforts in neural network design, Neural Architecture Search (NAS) has been applied with remarkable success to various high-level vision tasks such as classification and semantic segmentation. The underlying idea for the NAS algorithm is straightforward, namely, to enable the network the ability to choose among a set of operations (e.g., convolution with different filter sizes), one is able to find an optimal architecture that is better adapted to the problem at hand. However, so far the success of NAS has not been enjoyed by low-level geometric vision tasks such as stereo matching. This is partly due to the fact that state-of-the-art deep stereo matching networks, designed by humans, are already sheer in size. Directly applying the NAS to such massive structures is computationally prohibitive based on the currently available mainstream computing resources. In this paper, we propose the first end-to-end hierarchical NAS framework for deep stereo matching by incorporating task-specific human knowledge into the neural architecture search framework. Specifically, following the gold standard pipeline for deep stereo matching (i.e., feature extraction -- feature volume construction and dense matching), we optimize the architectures of the entire pipeline jointly. Extensive experiments show that our searched network outperforms all state-of-the-art deep stereo matching architectures and is ranked at the top 1 accuracy on KITTI stereo 2012, 2015 and Middlebury benchmarks, as well as the top 1 on SceneFlow dataset with a substantial improvement on the size of the network and the speed of inference. The code is available at https://github.com/XuelianCheng/LEAStereo.
Previous monocular depth estimation methods take a single view and directly regress the expected results. Though recent advances are made by applying geometrically inspired loss functions during training, the inference procedure does not explicitly impose any geometrical constraint. Therefore these models purely rely on the quality of data and the effectiveness of learning to generalize. This either leads to suboptimal results or the demand of huge amount of expensive ground truth labelled data to generate reasonable results. In this paper, we show for the first time that the monocular depth estimation problem can be reformulated as two sub-problems, a view synthesis procedure followed by stereo matching, with two intriguing properties, namely i) geometrical constraints can be explicitly imposed during inference; ii) demand on labelled depth data can be greatly alleviated. We show that the whole pipeline can still be trained in an end-to-end fashion and this new formulation plays a critical role in advancing the performance. The resulting model outperforms all the previous monocular depth estimation methods as well as the stereo block matching method in the challenging KITTI dataset by only using a small number of real training data. The model also generalizes well to other monocular depth estimation benchmarks. We also discuss the implications and the advantages of solving monocular depth estimation using stereo methods.
Stereo matching is a key component of autonomous driving perception. Recent unsupervised stereo matching approaches have received adequate attention due to their advantage of not requiring disparity ground truth. These approaches, however, perform poorly near occlusions. To overcome this drawback, in this paper, we propose CoT-Stereo, a novel unsupervised stereo matching approach. Specifically, we adopt a co-teaching framework where two networks interactively teach each other about the occlusions in an unsupervised fashion, which greatly improves the robustness of unsupervised stereo matching. Extensive experiments on the KITTI Stereo benchmarks demonstrate the superior performance of CoT-Stereo over all other state-of-the-art unsupervised stereo matching approaches in terms of both accuracy and speed. Our project webpage is https://sites.google.com/view/cot-stereo.
Deep end-to-end learning based stereo matching methods have achieved great success as witnessed by the leaderboards across different benchmarking datasets (KITTI, Middlebury, ETH3D, etc). However, real scenarios not only require approaches to have state-of-the-art performance but also real-time speed and domain-across generalization, which cannot be satisfied by existing methods. In this paper, we propose MSMD-Net (Multi-Scale and Multi-Dimension) to construct multi-scale and multi-dimension cost volume. At the multi-scale level, we generate four 4D combination volumes at different scales and integrate them with an encoder-decoder process to predict an initial disparity estimation. At the multi-dimension level, we additionally construct a 3D warped correlation volume and use it to refine the initial disparity map with residual learning. These two dimensional cost volumes are complementary to each other and can boost the performance of disparity estimation. Additionally, we propose a switch training strategy to alleviate the overfitting issue appeared in the pre-training process and further improve the generalization ability and accuracy of final disparity estimation. Our proposed method was evaluated on several benchmark datasets and ranked first on KITTI 2012 leaderboard and second on KITTI 2015 leaderboard as of September 9. In addition, our method shows strong domain-across generalization and outperforms best prior work by a noteworthy margin with three or even five times faster speed. The code of MSMD-Net is available at https://github.com/gallenszl/MSMD-Net.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
