Learning to reconstruct depths in a single image by watching unlabeled videos via deep convolutional network (DCN) is attracting significant attention in recent years. In this paper, we introduce a surface normal representation for unsupervised depth estimation framework. Our estimated depths are constrained to be compatible with predicted normals, yielding more robust geometry results. Specifically, we formulate an edge-aware depth-normal consistency term, and solve it by constructing a depth-to-normal layer and a normal-to-depth layer inside of the DCN. The depth-to-normal layer takes estimated depths as input, and computes normal directions using cross production based on neighboring pixels. Then given the estimated normals, the normal-to-depth layer outputs a regularized depth map through local planar smoothness. Both layers are computed with awareness of edges inside the image to help address the issue of depth/normal discontinuity and preserve sharp edges. Finally, to train the network, we apply the photometric error and gradient smoothness for both depth and normal predictions. We conducted experiments on both outdoor (KITTI) and indoor (NYUv2) datasets, and show that our algorithm vastly outperforms state of the art, which demonstrates the benefits from our approach.
Recently, coded masks have been used to demonstrate a thin form-factor lensless camera, FlatCam, in which a mask is placed immediately on top of a bare image sensor. In this paper, we present an imaging model and algorithm to jointly estimate depth and intensity information in the scene from a single or multiple FlatCams. We use a light field representation to model the mapping of 3D scene onto the sensor in which light rays from different depths yield different modulation patterns. We present a greedy depth pursuit algorithm to search the 3D volume and estimate the depth and intensity of each pixel within the camera field-of-view. We present simulation results to analyze the performance of our proposed model and algorithm with different FlatCam settings.
Many cameras implement auto-focus functionality. However, they typically require the user to manually identify the location to be focused on. While such an approach works for temporally-sparse autofocusing functionality (e.g., photo shooting), it presents extreme usability problems when the focus must be quickly switched between multiple areas (and depths) of interest - e.g., in a gaze-based autofocus approach. This work introduces a novel, real-time auto-focus approach based on eye-tracking, which enables the user to shift the camera focus plane swiftly based solely on the gaze information. Moreover, the proposed approach builds a graph representation of the image to estimate depth plane surfaces and runs in real time (requiring ~20ms on a single i5 core), thus allowing for the depth map estimation to be performed dynamically. We evaluated our algorithm for gaze-based depth estimation against state-of-the-art approaches based on eight new data sets with flat, skewed, and round surfaces, as well as publicly available datasets.
The paper presents a novel approach to occlusion handling problem in depth estimation using three views. A solution based on modification of similarity cost function is proposed. During the depth estimation via optimization algorithms like Graph Cut similarity metric is constantly updated so that only non-occluded fragments in side views are considered. At each iteration of the algorithm non-occluded fragments are detected based on side view virtual depth maps synthesized from the best currently estimated depth map of the center view. Then similarity metric is updated for correspondence search only in non-occluded regions of the side views. The experimental results, conducted on well-known 3D video test sequences, have proved that the depth maps estimated with the proposed approach provide about 1.25 dB virtual view quality improvement in comparison to the virtual view synthesized based on depth maps generated by the state-of-the-art MPEG Depth Estimation Reference Software.
Visual SLAM (Simultaneous Localization and Mapping) methods typically rely on handcrafted visual features or raw RGB values for establishing correspondences between images. These features, while suitable for sparse mapping, often lead to ambiguous matches in texture-less regions when performing dense reconstruction due to the aperture problem. In this work, we explore the use of learned features for the matching task in dense monocular reconstruction. We propose a novel convolutional neural network (CNN) architecture along with a deeply supervised feature learning scheme for pixel-wise regression of visual descriptors from an image which are best suited for dense monocular SLAM. In particular, our learning scheme minimizes a multi-view matching cost-volume loss with respect to the regressed features at multiple stages within the network, for explicitly learning contextual features that are suitable for dense matching between images captured by a moving monocular camera along the epipolar line. We integrate the learned features from our model for depth estimation inside a real-time dense monocular SLAM framework, where photometric error is replaced by our learned descriptor error. Our extensive evaluation on several challenging indoor datasets demonstrate greatly improved accuracy in dense reconstructions of the well celebrated dense SLAM systems like DTAM, without compromising their real-time performance.
Researches in novel viewpoint synthesis majorly focus on interpolation from multi-view input images. In this paper, we focus on a more challenging and ill-posed problem that is to synthesize novel viewpoints from one single input image. To achieve this goal, we propose a novel deep learning-based technique. We design a full resolution network that extracts local image features with the same resolution of the input, which contributes to derive high resolution and prevent blurry artifacts in the final synthesized images. We also involve a pre-trained depth estimation network into our system, and thus 3D information is able to be utilized to infer the flow field between the input and the target image. Since the depth network is trained by depth order information between arbitrary pairs of points in the scene, global image features are also involved into our system. Finally, a synthesis layer is used to not only warp the observed pixels to the desired positions but also hallucinate the missing pixels with recorded pixels. Experiments show that our technique performs well on images of various scenes, and outperforms the state-of-the-art techniques.
Multi-Camera arrays are increasingly employed in both consumer and industrial applications, and various passive techniques are documented to estimate depth from such camera arrays. Current depth estimation methods provide useful estimations of depth in an imaged scene but are often impractical due to significant computational requirements. This paper presents a novel framework that generates a high-quality continuous depth map from multi-camera array/light field cameras. The proposed framework utilizes analysis of the local Epipolar Plane Image (EPI) to initiate the depth estimation process. The estimated depth map is then processed using Total Variation (TV) minimization based on the Fenchel-Rockafellar duality. Evaluation of this method based on a well-known benchmark indicates that the proposed framework performs well in terms of accuracy when compared to the top-ranked depth estimation methods and a baseline algorithm. The test dataset includes both photorealistic and non-photorealistic scenes. Notably, the computational requirements required to achieve an equivalent accuracy are significantly reduced when compared to the top algorithms. As a consequence, the proposed framework is suitable for deployment in consumer and industrial applications.
Depth from focus (DFF) is one of the classical ill-posed inverse problems in computer vision. Most approaches recover the depth at each pixel based on the focal setting which exhibits maximal sharpness. Yet, it is not obvious how to reliably estimate the sharpness level, particularly in low-textured areas. In this paper, we propose `Deep Depth From Focus (DDFF)' as the first end-to-end learning approach to this problem. One of the main challenges we face is the hunger for data of deep neural networks. In order to obtain a significant amount of focal stacks with corresponding groundtruth depth, we propose to leverage a light-field camera with a co-calibrated RGB-D sensor. This allows us to digitally create focal stacks of varying sizes. Compared to existing benchmarks our dataset is 25 times larger, enabling the use of machine learning for this inverse problem. We compare our results with state-of-the-art DFF methods and we also analyze the effect of several key deep architectural components. These experiments show that our proposed method `DDFFNet' achieves state-of-the-art performance in all scenes, reducing depth error by more than 75% compared to the classical DFF methods.
Stereo depth estimation is error-prone; hence, effective error detection methods are desirable. Most such existing methods depend on characteristics of the stereo matching cost curve, making them unduly dependent on functional details of the matching algorithm. As a remedy, we propose a novel error detection approach based solely on the input image and its depth map. Our assumption is that, entropy of any point on an image will be significantly higher than the entropy of its corresponding point on the image's depth map. In this paper, we propose a confidence measure, Entropy-Difference (ED) for stereo depth estimates and a binary classification method to identify incorrect depths. Experiments on the Middlebury dataset show the effectiveness of our method. Our proposed stereo confidence measure outperforms 17 existing measures in all aspects except occlusion detection. Established metrics such as precision, accuracy, recall, and area-under-curve are used to demonstrate the effectiveness of our method.
Colorectal cancer is the fourth leading cause of cancer deaths worldwide and the second leading cause in the United States. The risk of colorectal cancer can be mitigated by the identification and removal of premalignant lesions through optical colonoscopy. Unfortunately, conventional colonoscopy misses more than 20% of the polyps that should be removed, due in part to poor contrast of lesion topography. Imaging tissue topography during a colonoscopy is difficult because of the size constraints of the endoscope and the deforming mucosa. Most existing methods make geometric assumptions or incorporate a priori information, which limits accuracy and sensitivity. In this paper, we present a method that avoids these restrictions, using a joint deep convolutional neural network-conditional random field (CNN-CRF) framework. Estimated depth is used to reconstruct the topography of the surface of the colon from a single image. We train the unary and pairwise potential functions of a CRF in a CNN on synthetic data, generated by developing an endoscope camera model and rendering over 100,000 images of an anatomically-realistic colon. We validate our approach with real endoscopy images from a porcine colon, transferred to a synthetic-like domain, with ground truth from registered computed tomography measurements. The CNN-CRF approach estimates depths with a relative error of 0.152 for synthetic endoscopy images and 0.242 for real endoscopy images. We show that the estimated depth maps can be used for reconstructing the topography of the mucosa from conventional colonoscopy images. This approach can easily be integrated into existing endoscopy systems and provides a foundation for improving computer-aided detection algorithms for detection, segmentation and classification of lesions.
To realize the full potential of deep learning for medical imaging, large annotated datasets are required for training. Such datasets are difficult to acquire because labeled medical images are not usually available due to privacy issues, lack of experts available for annotation, underrepresentation of rare conditions and poor standardization. Lack of annotated data has been addressed in conventional vision applications using synthetic images refined via unsupervised adversarial training to look like real images. However, this approach is difficult to extend to general medical imaging because of the complex and diverse set of features found in real human tissues. We propose an alternative framework that uses a reverse flow, where adversarial training is used to make real medical images more like synthetic images, and hypothesize that clinically-relevant features can be preserved via self-regularization. These domain-adapted images can then be accurately interpreted by networks trained on large datasets of synthetic medical images. We test this approach for the notoriously difficult task of depth-estimation from endoscopy. We train a depth estimator on a large dataset of synthetic images generated using an accurate forward model of an endoscope and an anatomically-realistic colon. This network predicts significantly better depths when using synthetic-like domain-adapted images compared to the real images, confirming that the clinically-relevant features of depth are preserved.
Many applications such as autonomous navigation, urban planning and asset monitoring, rely on the availability of accurate information about objects and their geolocations. In this paper we propose to automatically detect and compute the GPS coordinates of recurring stationary objects of interest using street view imagery. Our processing pipeline relies on two fully convolutional neural networks: the first segments objects in the images while the second estimates their distance from the camera. To geolocate all the detected objects coherently we propose a novel custom Markov Random Field model to perform objects triangulation. The novelty of the resulting pipeline is the combined use of monocular depth estimation and triangulation to enable automatic mapping of complex scenes with multiple visually similar objects of interest. We validate experimentally the effectiveness of our approach on two object classes: traffic lights and telegraph poles. The experiments report high object recall rates and GPS accuracy within 2 meters, which is comparable with the precision of single-frequency GPS receivers.
Objects may appear at arbitrary scales in perspective images of a scene, posing a challenge for recognition systems that process images at a fixed resolution. We propose a depth-aware gating module that adaptively selects the pooling field size in a convolutional network architecture according to the object scale (inversely proportional to the depth) so that small details are preserved for distant objects while larger receptive fields are used for those nearby. The depth gating signal is provided by stereo disparity or estimated directly from monocular input. We integrate this depth-aware gating into a recurrent convolutional neural network to perform semantic segmentation. Our recurrent module iteratively refines the segmentation results, leveraging the depth and semantic predictions from the previous iterations. Through extensive experiments on four popular large-scale RGB-D datasets, we demonstrate this approach achieves competitive semantic segmentation performance with a model which is substantially more compact. We carry out extensive analysis of this architecture including variants that operate on monocular RGB but use depth as side-information during training, unsupervised gating as a generic attentional mechanism, and multi-resolution gating. We find that gated pooling for joint semantic segmentation and depth yields state-of-the-art results for quantitative monocular depth estimation.
In this work, we propose an end-to-end deep architecture that jointly learns to detect obstacles and estimate their depth for MAV flight applications. Most of the existing approaches either rely on Visual SLAM systems or on depth estimation models to build 3D maps and detect obstacles. However, for the task of avoiding obstacles this level of complexity is not required. Recent works have proposed multi task architectures to both perform scene understanding and depth estimation. We follow their track and propose a specific architecture to jointly estimate depth and obstacles, without the need to compute a global map, but maintaining compatibility with a global SLAM system if needed. The network architecture is devised to exploit the joint information of the obstacle detection task, that produces more reliable bounding boxes, with the depth estimation one, increasing the robustness of both to scenario changes. We call this architecture J-MOD$^{2}$. We test the effectiveness of our approach with experiments on sequences with different appearance and focal lengths and compare it to SotA multi task methods that jointly perform semantic segmentation and depth estimation. In addition, we show the integration in a full system using a set of simulated navigation experiments where a MAV explores an unknown scenario and plans safe trajectories by using our detection model.
In this paper we consider the problem of single monocular image depth estimation. It is a challenging problem due to its ill-posedness nature and has found wide application in industry. Previous efforts belongs roughly to two families: learning-based method and interactive method. Learning-based method, in which deep convolutional neural network (CNN) is widely used, can achieve good result. But they suffer low generalization ability and typically perform poorly for unfamiliar scenes. Besides, data-hungry nature for such method makes data aquisition expensive and time-consuming. Interactive method requires human annotation of depth which, however, is errorneous and of large variance. To overcome these problems, we propose a new perspective for single monocular image depth estimation problem: size to depth. Our method require sparse label for real-world size of object rather than raw depth. A Coarse depth map is then inferred following geometric relationships according to size labels. Then we refine the depth map by doing energy function optimization on conditional random field(CRF). We experimentally demonstrate that our method outperforms traditional depth-labeling methods and can produce satisfactory depth maps.
Machine learning models are susceptible to adversarial perturbations: small changes to input that can cause large changes in output. It is also demonstrated that there exist input-agnostic perturbations, called universal adversarial perturbations, which can change the inference of target model on most of the data samples. However, existing methods to craft universal perturbations are (i) task specific, (ii) require samples from the training data distribution, and (iii) perform complex optimizations. Additionally, because of the data dependence, fooling ability of the crafted perturbations is proportional to the available training data. In this paper, we present a novel, generalizable and data-free approaches for crafting universal adversarial perturbations. Independent of the underlying task, our objective achieves fooling via corrupting the extracted features at multiple layers. Therefore, the proposed objective is generalizable to craft image-agnostic perturbations across multiple vision tasks such as object recognition, semantic segmentation, and depth estimation. In the practical setting of black-box attack scenario (when the attacker does not have access to the target model and it's training data), we show that our objective outperforms the data dependent objectives to fool the learned models. Further, via exploiting simple priors related to the data distribution, our objective remarkably boosts the fooling ability of the crafted perturbations. Significant fooling rates achieved by our objective emphasize that the current deep learning models are now at an increased risk, since our objective generalizes across multiple tasks without the requirement of training data for crafting the perturbations. To encourage reproducible research, we have released the codes for our proposed algorithm.
3D face shape is more expressive and viewpoint-consistent than its 2D counterpart. However, 3D facial landmark localization in a single image is challenging due to the ambiguous nature of landmarks under 3D perspective. Existing approaches typically adopt a suboptimal two-step strategy, performing 2D landmark localization followed by depth estimation. In this paper, we propose the Joint Voxel and Coordinate Regression (JVCR) method for 3D facial landmark localization, addressing it more effectively in an end-to-end fashion. First, a compact volumetric representation is proposed to encode the per-voxel likelihood of positions being the 3D landmarks. The dimensionality of such a representation is fixed regardless of the number of target landmarks, so that the curse of dimensionality could be avoided. Then, a stacked hourglass network is adopted to estimate the volumetric representation from coarse to fine, followed by a 3D convolution network that takes the estimated volume as input and regresses 3D coordinates of the face shape. In this way, the 3D structural constraints between landmarks could be learned by the neural network in a more efficient manner. Moreover, the proposed pipeline enables end-to-end training and improves the robustness and accuracy of 3D facial landmark localization. The effectiveness of our approach is validated on the 3DFAW and AFLW2000-3D datasets. Experimental results show that the proposed method achieves state-of-the-art performance in comparison with existing methods.
Digital holography enables us to reconstruct objects in three-dimensional space from holograms captured by an imaging device. For the reconstruction, we need to know the depth position of the recoded object in advance. In this study, we propose depth prediction using convolutional neural network (CNN)-based regression. In the previous researches, the depth of an object was estimated through reconstructed images at different depth positions from a hologram using a certain metric that indicates the most focused depth position; however, such a depth search is time-consuming. The CNN of the proposed method can directly predict the depth position with millimeter precision from holograms.
We present a generalization of the Cauchy/Lorentzian, Geman-McClure, Welsch/Leclerc, generalized Charbonnier, Charbonnier/pseudo-Huber/L1-L2, and L2 loss functions. By introducing robustness as a continuous parameter, our loss function allows algorithms built around robust loss minimization to be generalized, which improves performance on basic vision tasks such as registration and clustering. Interpreting our loss as the negative log of a univariate density yields a general probability distribution that includes normal and Cauchy distributions as special cases. This probabilistic interpretation enables the training of neural networks in which the robustness of the loss automatically adapts itself during training, which improves performance on learning-based tasks such as generative image synthesis and unsupervised monocular depth estimation, without requiring any manual parameter tuning.
Light field imaging extends the traditional photography by capturing both spatial and angular distribution of light, which enables new capabilities, including post-capture refocusing, post-capture aperture control, and depth estimation from a single shot. Micro-lens array (MLA) based light field cameras offer a cost-effective approach to capture light field. A major drawback of MLA based light field cameras is low spatial resolution, which is due to the fact that a single image sensor is shared to capture both spatial and angular information. In this paper, we present a learning based light field enhancement approach. Both spatial and angular resolution of captured light field is enhanced using convolutional neural networks. The proposed method is tested with real light field data captured with a Lytro light field camera, clearly demonstrating spatial and angular resolution improvement.