It is widely believed that sparse supervision is worse than dense supervision in the field of depth completion, but the underlying reasons for this are rarely discussed. To this end, we revisit the task of radar-camera depth completion and present a new method with sparse LiDAR supervision to outperform previous dense LiDAR supervision methods in both accuracy and speed. Specifically, when trained by sparse LiDAR supervision, depth completion models usually output depth maps containing significant stripe-like artifacts. We find that such a phenomenon is caused by the implicitly learned positional distribution pattern from sparse LiDAR supervision, termed as LiDAR Distribution Leakage (LDL) in this paper. Based on such understanding, we present a novel Disruption-Compensation radar-camera depth completion framework to address this issue. The Disruption part aims to deliberately disrupt the learning of LiDAR distribution from sparse supervision, while the Compensation part aims to leverage 3D spatial and 2D semantic information to compensate for the information loss of previous disruptions. Extensive experimental results demonstrate that by reducing the impact of LDL, our framework with sparse supervision outperforms the state-of-the-art dense supervision methods with 11.6% improvement in Mean Absolute Error (MAE)} and 1.6x speedup in Frame Per Second (FPS)}. The code is available at https://github.com/megvii-research/Sparse-Beats-Dense.
Colonoscopy screening is the gold standard procedure for assessing abnormalities in the colon and rectum, such as ulcers and cancerous polyps. Measuring the abnormal mucosal area and its 3D reconstruction can help quantify the surveyed area and objectively evaluate disease burden. However, due to the complex topology of these organs and variable physical conditions, for example, lighting, large homogeneous texture, and image modality estimating distance from the camera aka depth) is highly challenging. Moreover, most colonoscopic video acquisition is monocular, making the depth estimation a non-trivial problem. While methods in computer vision for depth estimation have been proposed and advanced on natural scene datasets, the efficacy of these techniques has not been widely quantified on colonoscopy datasets. As the colonic mucosa has several low-texture regions that are not well pronounced, learning representations from an auxiliary task can improve salient feature extraction, allowing estimation of accurate camera depths. In this work, we propose to develop a novel multi-task learning (MTL) approach with a shared encoder and two decoders, namely a surface normal decoder and a depth estimator decoder. Our depth estimator incorporates attention mechanisms to enhance global context awareness. We leverage the surface normal prediction to improve geometric feature extraction. Also, we apply a cross-task consistency loss among the two geometrically related tasks, surface normal and camera depth. We demonstrate an improvement of 14.17% on relative error and 10.4% improvement on $\delta_{1}$ accuracy over the most accurate baseline state-of-the-art BTS approach. All experiments are conducted on a recently released C3VD dataset; thus, we provide a first benchmark of state-of-the-art methods.
Despite the recent success of Neural Radiance Field (NeRF), it is still challenging to render large-scale driving scenes with long trajectories, particularly when the rendering quality and efficiency are in high demand. Existing methods for such scenes usually involve with spatial warping, geometric supervision from zero-shot normal or depth estimation, or scene division strategies, where the synthesized views are often blurry or fail to meet the requirement of efficient rendering. To address the above challenges, this paper presents a novel framework that learns a density space from the scenes to guide the construction of a point-based renderer, dubbed as DGNR (Density-Guided Neural Rendering). In DGNR, geometric priors are no longer needed, which can be intrinsically learned from the density space through volumetric rendering. Specifically, we make use of a differentiable renderer to synthesize images from the neural density features obtained from the learned density space. A density-based fusion module and geometric regularization are proposed to optimize the density space. By conducting experiments on a widely used autonomous driving dataset, we have validated the effectiveness of DGNR in synthesizing photorealistic driving scenes and achieving real-time capable rendering.
Multi-camera setups find widespread use across various applications, such as autonomous driving, as they greatly expand sensing capabilities. Despite the fast development of Neural radiance field (NeRF) techniques and their wide applications in both indoor and outdoor scenes, applying NeRF to multi-camera systems remains very challenging. This is primarily due to the inherent under-calibration issues in multi-camera setup, including inconsistent imaging effects stemming from separately calibrated image signal processing units in diverse cameras, and system errors arising from mechanical vibrations during driving that affect relative camera poses. In this paper, we present UC-NeRF, a novel method tailored for novel view synthesis in under-calibrated multi-view camera systems. Firstly, we propose a layer-based color correction to rectify the color inconsistency in different image regions. Second, we propose virtual warping to generate more viewpoint-diverse but color-consistent virtual views for color correction and 3D recovery. Finally, a spatiotemporally constrained pose refinement is designed for more robust and accurate pose calibration in multi-camera systems. Our method not only achieves state-of-the-art performance of novel view synthesis in multi-camera setups, but also effectively facilitates depth estimation in large-scale outdoor scenes with the synthesized novel views.
Modern learning-based approaches to 3D-aware image synthesis achieve high photorealism and 3D-consistent viewpoint changes for the generated images. Existing approaches represent instances in a shared canonical space. However, for in-the-wild datasets a shared canonical system can be difficult to define or might not even exist. In this work, we instead model instances in view space, alleviating the need for posed images and learned camera distributions. We find that in this setting, existing GAN-based methods are prone to generating flat geometry and struggle with distribution coverage. We hence propose WildFusion, a new approach to 3D-aware image synthesis based on latent diffusion models (LDMs). We first train an autoencoder that infers a compressed latent representation, which additionally captures the images' underlying 3D structure and enables not only reconstruction but also novel view synthesis. To learn a faithful 3D representation, we leverage cues from monocular depth prediction. Then, we train a diffusion model in the 3D-aware latent space, thereby enabling synthesis of high-quality 3D-consistent image samples, outperforming recent state-of-the-art GAN-based methods. Importantly, our 3D-aware LDM is trained without any direct supervision from multiview images or 3D geometry and does not require posed images or learned pose or camera distributions. It directly learns a 3D representation without relying on canonical camera coordinates. This opens up promising research avenues for scalable 3D-aware image synthesis and 3D content creation from in-the-wild image data. See https://katjaschwarz.github.io/wildfusion for videos of our 3D results.
In this paper, we present a method to optimize Gaussian splatting with a limited number of images while avoiding overfitting. Representing a 3D scene by combining numerous Gaussian splats has yielded outstanding visual quality. However, it tends to overfit the training views when only a small number of images are available. To address this issue, we introduce a dense depth map as a geometry guide to mitigate overfitting. We obtained the depth map using a pre-trained monocular depth estimation model and aligning the scale and offset using sparse COLMAP feature points. The adjusted depth aids in the color-based optimization of 3D Gaussian splatting, mitigating floating artifacts, and ensuring adherence to geometric constraints. We verify the proposed method on the NeRF-LLFF dataset with varying numbers of few images. Our approach demonstrates robust geometry compared to the original method that relies solely on images. Project page: robot0321.github.io/DepthRegGS
Monocular depth estimation is an important step in many downstream tasks in machine vision. We address the topic of estimating monocular depth from defocus blur which can yield more accurate results than the semantic based depth estimation methods. The existing monocular depth from defocus techniques are sensitive to the particular camera that the images are taken from. We show how several camera-related parameters affect the defocus blur using optical physics equations and how they make the defocus blur depend on these parameters. The simple correction procedure we propose can alleviate this problem which does not require any retraining of the original model. We created a synthetic dataset which can be used to test the camera independent performance of depth from defocus blur models. We evaluate our model on both synthetic and real datasets (DDFF12 and NYU depth V2) obtained with different cameras and show that our methods are significantly more robust to the changes of cameras. Code: https://github.com/sleekEagle/defocus_camind.git
Scene segmentation via unsupervised domain adaptation (UDA) enables the transfer of knowledge acquired from source synthetic data to real-world target data, which largely reduces the need for manual pixel-level annotations in the target domain. To facilitate domain-invariant feature learning, existing methods typically mix data from both the source domain and target domain by simply copying and pasting the pixels. Such vanilla methods are usually sub-optimal since they do not take into account how well the mixed layouts correspond to real-world scenarios. Real-world scenarios are with an inherent layout. We observe that semantic categories, such as sidewalks, buildings, and sky, display relatively consistent depth distributions, and could be clearly distinguished in a depth map. Based on such observation, we propose a depth-aware framework to explicitly leverage depth estimation to mix the categories and facilitate the two complementary tasks, i.e., segmentation and depth learning in an end-to-end manner. In particular, the framework contains a Depth-guided Contextual Filter (DCF) forndata augmentation and a cross-task encoder for contextual learning. DCF simulates the real-world layouts, while the cross-task encoder further adaptively fuses the complementing features between two tasks. Besides, it is worth noting that several public datasets do not provide depth annotation. Therefore, we leverage the off-the-shelf depth estimation network to generate the pseudo depth. Extensive experiments show that our proposed methods, even with pseudo depth, achieve competitive performance on two widely-used bench-marks, i.e. 77.7 mIoU on GTA to Cityscapes and 69.3 mIoU on Synthia to Cityscapes.
Adversarial attacks present a significant challenge to the dependable deployment of machine learning models, with patch-based attacks being particularly potent. These attacks introduce adversarial perturbations in localized regions of an image, deceiving even well-trained models. In this paper, we propose Outlier Detection and Dimension Reduction (ODDR), a comprehensive defense strategy engineered to counteract patch-based adversarial attacks through advanced statistical methodologies. Our approach is based on the observation that input features corresponding to adversarial patches-whether naturalistic or synthetic-deviate from the intrinsic distribution of the remaining image data and can thus be identified as outliers. ODDR operates through a robust three-stage pipeline: Fragmentation, Segregation, and Neutralization. This model-agnostic framework is versatile, offering protection across various tasks, including image classification, object detection, and depth estimation, and is proved effective in both CNN-based and Transformer-based architectures. In the Fragmentation stage, image samples are divided into smaller segments, preparing them for the Segregation stage, where advanced outlier detection techniques isolate anomalous features linked to adversarial perturbations. The Neutralization stage then applies dimension reduction techniques to these outliers, effectively neutralizing the adversarial impact while preserving critical information for the machine learning task. Extensive evaluation on benchmark datasets against state-of-the-art adversarial patches underscores the efficacy of ODDR. Our method enhances model accuracy from 39.26% to 79.1% under the GoogleAp attack, outperforming leading defenses such as LGS (53.86%), Jujutsu (60%), and Jedi (64.34%).
3D occupancy prediction is an important task for the robustness of vision-centric autonomous driving, which aims to predict whether each point is occupied in the surrounding 3D space. Existing methods usually require 3D occupancy labels to produce meaningful results. However, it is very laborious to annotate the occupancy status of each voxel. In this paper, we propose SelfOcc to explore a self-supervised way to learn 3D occupancy using only video sequences. We first transform the images into the 3D space (e.g., bird's eye view) to obtain 3D representation of the scene. We directly impose constraints on the 3D representations by treating them as signed distance fields. We can then render 2D images of previous and future frames as self-supervision signals to learn the 3D representations. We propose an MVS-embedded strategy to directly optimize the SDF-induced weights with multiple depth proposals. Our SelfOcc outperforms the previous best method SceneRF by 58.7% using a single frame as input on SemanticKITTI and is the first self-supervised work that produces reasonable 3D occupancy for surround cameras on nuScenes. SelfOcc produces high-quality depth and achieves state-of-the-art results on novel depth synthesis, monocular depth estimation, and surround-view depth estimation on the SemanticKITTI, KITTI-2015, and nuScenes, respectively. Code: https://github.com/huang-yh/SelfOcc.
Depth estimation from a single image is a challenging problem in computer vision because binocular disparity or motion information is absent. Whereas impressive performances have been reported in this area recently using end-to-end trained deep neural architectures, as to what cues in the images that are being exploited by these black box systems is hard to know. To this end, in this work, we quantify the relative contributions of the known cues of depth in a monocular depth estimation setting using an indoor scene data set. Our work uses feature extraction techniques to relate the single features of shape, texture, colour and saturation, taken in isolation, to predict depth. We find that the shape of objects extracted by edge detection substantially contributes more than others in the indoor setting considered, while the other features also have contributions in varying degrees. These insights will help optimise depth estimation models, boosting their accuracy and robustness. They promise to broaden the practical applications of vision-based depth estimation. The project code is attached to the supplementary material and will be published on GitHub.
Inverse rendering is an ill-posed problem. Previous work has sought to resolve this by focussing on priors for object or scene shape or appearance. In this work, we instead focus on a prior for natural illuminations. Current methods rely on spherical harmonic lighting or other generic representations and, at best, a simplistic prior on the parameters. This results in limitations for the inverse setting in terms of the expressivity of the illumination conditions, especially when taking specular reflections into account. We propose a conditional neural field representation based on a variational auto-decoder and a transformer decoder. We extend Vector Neurons to build equivariance directly into our architecture, and leveraging insights from depth estimation through a scale-invariant loss function, we enable the accurate representation of High Dynamic Range (HDR) images. The result is a compact, rotation-equivariant HDR neural illumination model capable of capturing complex, high-frequency features in natural environment maps. Training our model on a curated dataset of 1.6K HDR environment maps of natural scenes, we compare it against traditional representations, demonstrate its applicability for an inverse rendering task and show environment map completion from partial observations. We share our PyTorch implementation, dataset and trained models at https://github.com/JADGardner/ns_reni
Personal autonomous vehicles are cars, trucks and bikes capable of sensing their surrounding environment, planning their route, and driving with little or no involvement of human drivers. Despite the impressive technological achievements made by the industry in recent times and the hopeful announcements made by leading entrepreneurs, to date no personal vehicle is approved for road circulation in a 'fully' or 'semi' autonomous mode (autonomy levels 4 and 5) and it is still unclear when such vehicles will eventually be mature enough to receive this kind of approval. The present review adopts an integrative and multidisciplinary approach to investigate the major challenges faced by the automative sector, with the aim to identify the problems that still trouble and delay the commercialization of autonomous vehicles. The review examines the limitations and risks associated with current technologies and the most promising solutions devised by the researchers. This negative assessment methodology is not motivated by pessimism, but by the aspiration to raise critical awareness about the technology's state-of-the-art, the industry's quality standards, and the society's demands and expectations. While the survey primarily focuses on the applications of artificial intelligence for perception and navigation, it also aims to offer an enlarged picture that links the purely technological aspects with the relevant human-centric aspects, including, cultural attitudes, conceptual assumptions, and normative (ethico-legal) frameworks. Examining the broader context serves to highlight problems that have a cross-disciplinary scope and identify solutions that may benefit from a holistic consideration.
Light field cameras have a wide range of uses due to their ability to simultaneously record light intensity and direction. The angular resolution of light fields is important for downstream tasks such as depth estimation, yet is often difficult to improve due to hardware limitations. Conventional methods tend to perform poorly against the challenge of large disparity in sparse light fields, while general CNNs have difficulty extracting spatial and angular features coupled together in 4D light fields. The light field disentangling mechanism transforms the 4D light field into 2D image format, which is more favorable for CNN for feature extraction. In this paper, we propose a Deep Disentangling Mechanism, which inherits the principle of the light field disentangling mechanism and further develops the design of the feature extractor and adds advanced network structure. We design a light-field reconstruction network (i.e., DDASR) on the basis of the Deep Disentangling Mechanism, and achieve SOTA performance in the experiments. In addition, we design a Block Traversal Angular Super-Resolution Strategy for the practical application of depth estimation enhancement where the input views is often higher than 2x2 in the experiments resulting in a high memory usage, which can reduce the memory usage while having a better reconstruction performance.
Despite the enhanced realism and immersion provided by VR headsets, users frequently encounter adverse effects such as digital eye strain (DES), dry eye, and potential long-term visual impairment due to excessive eye stimulation from VR displays and pressure from the mask. Recent VR headsets are increasingly equipped with eye-oriented monocular cameras to segment ocular feature maps. Yet, to compute the incident light stimulus and observe periocular condition alterations, it is imperative to transform these relative measurements into metric dimensions. To bridge this gap, we propose a lightweight framework derived from the U-Net 3+ deep learning backbone that we re-optimised, to estimate measurable periocular depth maps. Compatible with any VR headset equipped with an eye-oriented monocular camera, our method reconstructs three-dimensional periocular regions, providing a metric basis for related light stimulus calculation protocols and medical guidelines. Navigating the complexities of data collection, we introduce a Dynamic Periocular Data Generation (DPDG) environment based on UE MetaHuman, which synthesises thousands of training images from a small quantity of human facial scan data. Evaluated on a sample of 36 participants, our method exhibited notable efficacy in the periocular global precision evaluation experiment, and the pupil diameter measurement.
Over the past few years, self-supervised monocular depth estimation that does not depend on ground-truth during the training phase has received widespread attention. Most efforts focus on designing different types of network architectures and loss functions or handling edge cases, e.g., occlusion and dynamic objects. In this work, we introduce a novel self-supervised depth estimation framework, dubbed MonoDiffusion, by formulating it as an iterative denoising process. Because the depth ground-truth is unavailable in the training phase, we develop a pseudo ground-truth diffusion process to assist the diffusion in MonoDiffusion. The pseudo ground-truth diffusion gradually adds noise to the depth map generated by a pre-trained teacher model. Moreover,the teacher model allows applying a distillation loss to guide the denoised depth. Further, we develop a masked visual condition mechanism to enhance the denoising ability of model. Extensive experiments are conducted on the KITTI and Make3D datasets and the proposed MonoDiffusion outperforms prior state-of-the-art competitors. The source code will be available at https://github.com/ShuweiShao/MonoDiffusion.
Over the past few years, monocular depth estimation and completion have been paid more and more attention from the computer vision community because of their widespread applications. In this paper, we introduce novel physics (geometry)-driven deep learning frameworks for these two tasks by assuming that 3D scenes are constituted with piece-wise planes. Instead of directly estimating the depth map or completing the sparse depth map, we propose to estimate the surface normal and plane-to-origin distance maps or complete the sparse surface normal and distance maps as intermediate outputs. To this end, we develop a normal-distance head that outputs pixel-level surface normal and distance. Meanwhile, the surface normal and distance maps are regularized by a developed plane-aware consistency constraint, which are then transformed into depth maps. Furthermore, we integrate an additional depth head to strengthen the robustness of the proposed frameworks. Extensive experiments on the NYU-Depth-v2, KITTI and SUN RGB-D datasets demonstrate that our method exceeds in performance prior state-of-the-art monocular depth estimation and completion competitors. The source code will be available at https://github.com/ShuweiShao/NDDepth.
Self-supervised monocular depth estimation methods aim to be used in critical applications such as autonomous vehicles for environment analysis. To circumvent the potential imperfections of these approaches, a quantification of the prediction confidence is crucial to guide decision-making systems that rely on depth estimation. In this paper, we propose MonoProb, a new unsupervised monocular depth estimation method that returns an interpretable uncertainty, which means that the uncertainty reflects the expected error of the network in its depth predictions. We rethink the stereo or the structure-from-motion paradigms used to train unsupervised monocular depth models as a probabilistic problem. Within a single forward pass inference, this model provides a depth prediction and a measure of its confidence, without increasing the inference time. We then improve the performance on depth and uncertainty with a novel self-distillation loss for which a student is supervised by a pseudo ground truth that is a probability distribution on depth output by a teacher. To quantify the performance of our models we design new metrics that, unlike traditional ones, measure the absolute performance of uncertainty predictions. Our experiments highlight enhancements achieved by our method on standard depth and uncertainty metrics as well as on our tailored metrics. https://github.com/CEA-LIST/MonoProb
Dense prediction tasks, such as semantic segmentation, depth estimation, and surface normal prediction, can be easily formulated as per-pixel classification (discrete outputs) or regression (continuous outputs). This per-pixel prediction paradigm has remained popular due to the prevalence of fully convolutional networks. However, on the recent frontier of segmentation task, the community has been witnessing a shift of paradigm from per-pixel prediction to cluster-prediction with the emergence of transformer architectures, particularly the mask transformers, which directly predicts a label for a mask instead of a pixel. Despite this shift, methods based on the per-pixel prediction paradigm still dominate the benchmarks on the other dense prediction tasks that require continuous outputs, such as depth estimation and surface normal prediction. Motivated by the success of DORN and AdaBins in depth estimation, achieved by discretizing the continuous output space, we propose to generalize the cluster-prediction based method to general dense prediction tasks. This allows us to unify dense prediction tasks with the mask transformer framework. Remarkably, the resulting model PolyMaX demonstrates state-of-the-art performance on three benchmarks of NYUD-v2 dataset. We hope our simple yet effective design can inspire more research on exploiting mask transformers for more dense prediction tasks. Code and model will be made available.
Colonoscopy is the choice procedure to diagnose colon and rectum cancer, from early detection of small precancerous lesions (polyps), to confirmation of malign masses. However, the high variability of the organ appearance and the complex shape of both the colon wall and structures of interest make this exploration difficult. Learned visuospatial and perceptual abilities mitigate technical limitations in clinical practice by proper estimation of the intestinal depth. This work introduces a novel methodology to estimate colon depth maps in single frames from monocular colonoscopy videos. The generated depth map is inferred from the shading variation of the colon wall with respect to the light source, as learned from a realistic synthetic database. Briefly, a classic convolutional neural network architecture is trained from scratch to estimate the depth map, improving sharp depth estimations in haustral folds and polyps by a custom loss function that minimizes the estimation error in edges and curvatures. The network was trained by a custom synthetic colonoscopy database herein constructed and released, composed of 248,400 frames (47 videos), with depth annotations at the level of pixels. This collection comprehends 5 subsets of videos with progressively higher levels of visual complexity. Evaluation of the depth estimation with the synthetic database reached a threshold accuracy of 95.65%, and a mean-RMSE of 0.451 cm, while a qualitative assessment with a real database showed consistent depth estimations, visually evaluated by the expert gastroenterologist coauthoring this paper. Finally, the method achieved competitive performance with respect to another state-of-the-art method using a public synthetic database and comparable results in a set of images with other five state-of-the-art methods.