Katherine A. Skinner

Alexandra Carlson, Manikandasriram Srinivasan Ramanagopal, Nathan Tseng et al.

Recent advances in neural radiance fields (NeRFs) achieve state-of-the-art novel view synthesis and facilitate dense estimation of scene properties. However, NeRFs often fail for large, unbounded scenes that are captured under very sparse views with the scene content concentrated far away from the camera, as is typical for field robotics applications. In particular, NeRF-style algorithms perform poorly: (1) when there are insufficient views with little pose diversity, (2) when scenes contain saturation and shadows, and (3) when finely sampling large unbounded scenes with fine structures becomes computationally intensive. This paper proposes CLONeR, which significantly improves upon NeRF by allowing it to model large outdoor driving scenes that are observed from sparse input sensor views. This is achieved by decoupling occupancy and color learning within the NeRF framework into separate Multi-Layer Perceptrons (MLPs) trained using LiDAR and camera data, respectively. In addition, this paper proposes a novel method to build differentiable 3D Occupancy Grid Maps (OGM) alongside the NeRF model, and leverage this occupancy grid for improved sampling of points along a ray for volumetric rendering in metric space. Through extensive quantitative and qualitative experiments on scenes from the KITTI dataset, this paper demonstrates that the proposed method outperforms state-of-the-art NeRF models on both novel view synthesis and dense depth prediction tasks when trained on sparse input data.

Jingyu Song, Xudong Chen, Liupei Lu et al.

High-definition (HD) maps provide environmental information for autonomous driving systems and are essential for safe planning. While existing methods with single-frame input achieve impressive performance for online vectorized HD map construction, they still struggle with complex scenarios and occlusions. We propose MemFusionMap, a novel temporal fusion model with enhanced temporal reasoning capabilities for online HD map construction. Specifically, we contribute a working memory fusion module that improves the model's memory capacity to reason across a history of frames. We also design a novel temporal overlap heatmap to explicitly inform the model about the temporal overlap information and vehicle trajectory in the Bird's Eye View space. By integrating these two designs, MemFusionMap significantly outperforms existing methods while also maintaining a versatile design for scalability. We conduct extensive evaluation on open-source benchmarks and demonstrate a maximum improvement of 5.4% in mAP over state-of-the-art methods. The project page for MemFusionMap is https://song-jingyu.github.io/MemFusionMap

Advaith Venkatramanan Sethuraman, Manikandasriram Srinivasan Ramanagopal, Katherine A. Skinner

Underwater imaging is a critical task performed by marine robots for a wide range of applications including aquaculture, marine infrastructure inspection, and environmental monitoring. However, water column effects, such as attenuation and backscattering, drastically change the color and quality of imagery captured underwater. Due to varying water conditions and range-dependency of these effects, restoring underwater imagery is a challenging problem. This impacts downstream perception tasks including depth estimation and 3D reconstruction. In this paper, we advance state-of-the-art in neural radiance fields (NeRFs) to enable physics-informed dense depth estimation and color correction. Our proposed method, WaterNeRF, estimates parameters of a physics-based model for underwater image formation, leading to a hybrid data-driven and model-based solution. After determining the scene structure and radiance field, we can produce novel views of degraded as well as corrected underwater images, along with dense depth of the scene. We evaluate the proposed method qualitatively and quantitatively on a real underwater dataset.

Anja Sheppard, Parker Ewen, Joey Wilson et al.

This paper introduces SLIM-VDB, a new lightweight semantic mapping system with probabilistic semantic fusion for closed-set or open-set dictionaries. Advances in data structures from the computer graphics community, such as OpenVDB, have demonstrated significantly improved computational and memory efficiency in volumetric scene representation. Although OpenVDB has been used for geometric mapping in robotics applications, semantic mapping for scene understanding with OpenVDB remains unexplored. In addition, existing semantic mapping systems lack support for integrating both fixed-category and open-language label predictions within a single framework. In this paper, we propose a novel 3D semantic mapping system that leverages the OpenVDB data structure and integrates a unified Bayesian update framework for both closed- and open-set semantic fusion. Our proposed framework, SLIM-VDB, achieves significant reduction in both memory and integration times compared to current state-of-the-art semantic mapping approaches, while maintaining comparable mapping accuracy. An open-source C++ codebase with a Python interface is available at https://github.com/umfieldrobotics/slim-vdb.

Advaith Venkatramanan Sethuraman, Katherine A. Skinner

In this paper, we address the problem of sim-to-real transfer for object segmentation when there is no access to real examples of an object of interest during training, i.e. zero-shot sim-to-real transfer for segmentation. We focus on the application of shipwreck segmentation in side scan sonar imagery. Our novel segmentation network, STARS, addresses this challenge by fusing a predicted deformation field and anomaly volume, allowing it to generalize better to real sonar images and achieve more effective zero-shot sim-to-real transfer for image segmentation. We evaluate the sim-to-real transfer capabilities of our method on a real, expert-labeled side scan sonar dataset of shipwrecks collected from field work surveys with an autonomous underwater vehicle (AUV). STARS is trained entirely in simulation and performs zero-shot shipwreck segmentation with no additional fine-tuning on real data. Our method provides a significant 20% increase in segmentation performance for the targeted shipwreck class compared to the best baseline.

Spencer Carmichael, Katherine A. Skinner

Thermal cameras offer several advantages for simultaneous localization and mapping (SLAM) with mobile robots: they provide a passive, low-power solution to operating in darkness, are invariant to rapidly changing or high dynamic range illumination, and can see through fog, dust, and smoke. However, uncooled microbolometer thermal cameras, the only practical option in most robotics applications, suffer from significant motion blur, rolling shutter distortions, and fixed pattern noise. In this paper, we present TRGS-SLAM, a 3D Gaussian Splatting (3DGS) based thermal inertial SLAM system uniquely capable of handling these degradations. To overcome the challenges of thermal data, we introduce a model-aware 3DGS rendering method and several general innovations to 3DGS SLAM, including B-spline trajectory optimization with a two-stage IMU loss, view-diversity-based opacity resetting, and pose drift correction schemes. Our system demonstrates accurate tracking on real-world, fast motion, and high-noise thermal data that causes all other tested SLAM methods to fail. Moreover, through offline refinement of our SLAM results, we demonstrate thermal image restoration competitive with prior work that required ground truth poses.

Anja Sheppard, Tyler Smithline, Andrew Scheffer et al.

In this paper, we introduce ShipwreckFinder, an open-source QGIS plugin that detects shipwrecks from multibeam sonar data. Shipwrecks are an important historical marker of maritime history, and can be discovered through manual inspection of bathymetric data. However, this is a time-consuming process and often requires expert analysis. Our proposed tool allows users to automatically preprocess bathymetry data, perform deep learning inference, threshold model outputs, and produce either pixel-wise segmentation masks or bounding boxes of predicted shipwrecks. The backbone of this open-source tool is a deep learning model, which is trained on a variety of shipwreck data from the Great Lakes and the coasts of Ireland. Additionally, we employ synthetic data generation in order to increase the size and diversity of our dataset. We demonstrate superior segmentation performance with our open-source tool and training pipeline as compared to a deep learning-based ArcGIS toolkit and a more classical inverse sinkhole detection method. The open-source tool can be found at https://github.com/umfieldrobotics/ShipwreckFinderQGISPlugin.

Yeheng Zong, Pou-Chun Kung, Yike Pan et al.

Accurately recovering human pose and appearance from video is an essential component of scene reconstruction, with applications to motion capture, motion prediction, virtual reality, and digital twinning. Despite significant interest in building realistic human avatars from video, this paper demonstrates that existing methods do not accurately recover the 3D geometry of humans. ViT-based approaches are not consistently reliable and can overfit to 2D views, while NeRF- and Gaussian Splatting-based avatars treat pose and appearance separately, limiting rendering generalization to new poses. To resolve these shortcomings, this paper proposes HumanSplatHMR, a joint optimization framework that refines 3D human poses while simultaneously learning a high-fidelity avatar for novel-view and novel-pose synthesis. Our key insight is to close the loop between geometric pose estimation and differentiable rendering. Unlike prior human avatar methods that rely on accurate human pose obtained through motion capture systems or offline refinement, which are impractical in in-the-wild scenarios, our approach uses only human mesh estimates from a state-of-the-art human pose estimator to better reflect real-world conditions. Therefore, instead of using the human pose only as a deformation prior, HumanSplatHMR backpropagates photometric, segmentation, and depth losses through a differentiable renderer to the pose parameters and global position. This coupling refines the global 3D pose over time, improving accuracy and alignment while producing better renderings from novel views. Experiments show consistent improvements over pose recovery baselines that omit image-level refinement and avatar baselines that decouple pose estimation from avatar reconstruction.

Lingjun Zhao, Jingyu Song, Katherine A. Skinner

In the field of 3D object detection for autonomous driving, LiDAR-Camera (LC) fusion is the top-performing sensor configuration. Still, LiDAR is relatively high cost, which hinders adoption of this technology for consumer automobiles. Alternatively, camera and radar are commonly deployed on vehicles already on the road today, but performance of Camera-Radar (CR) fusion falls behind LC fusion. In this work, we propose Camera-Radar Knowledge Distillation (CRKD) to bridge the performance gap between LC and CR detectors with a novel cross-modality KD framework. We use the Bird's-Eye-View (BEV) representation as the shared feature space to enable effective knowledge distillation. To accommodate the unique cross-modality KD path, we propose four distillation losses to help the student learn crucial features from the teacher model. We present extensive evaluations on the nuScenes dataset to demonstrate the effectiveness of the proposed CRKD framework. The project page for CRKD is https://song-jingyu.github.io/CRKD.

Jingyu Song, Lingjun Zhao, Katherine A. Skinner

We propose LiRaFusion to tackle LiDAR-radar fusion for 3D object detection to fill the performance gap of existing LiDAR-radar detectors. To improve the feature extraction capabilities from these two modalities, we design an early fusion module for joint voxel feature encoding, and a middle fusion module to adaptively fuse feature maps via a gated network. We perform extensive evaluation on nuScenes to demonstrate that LiRaFusion leverages the complementary information of LiDAR and radar effectively and achieves notable improvement over existing methods.

Pou-Chun Kung, Xianling Zhang, Katherine A. Skinner et al.

Photorealistic 3D scene reconstruction plays an important role in autonomous driving, enabling the generation of novel data from existing datasets to simulate safety-critical scenarios and expand training data without additional acquisition costs. Gaussian Splatting (GS) facilitates real-time, photorealistic rendering with an explicit 3D Gaussian representation of the scene, providing faster processing and more intuitive scene editing than the implicit Neural Radiance Fields (NeRFs). While extensive GS research has yielded promising advancements in autonomous driving applications, they overlook two critical aspects: First, existing methods mainly focus on low-speed and feature-rich urban scenes and ignore the fact that highway scenarios play a significant role in autonomous driving. Second, while LiDARs are commonplace in autonomous driving platforms, existing methods learn primarily from images and use LiDAR only for initial estimates or without precise sensor modeling, thus missing out on leveraging the rich depth information LiDAR offers and limiting the ability to synthesize LiDAR data. In this paper, we propose a novel GS method for dynamic scene synthesis and editing with improved scene reconstruction through LiDAR supervision and support for LiDAR rendering. Unlike prior works that are tested mostly on urban datasets, to the best of our knowledge, we are the first to focus on the more challenging and highly relevant highway scenes for autonomous driving, with sparse sensor views and monotone backgrounds. Visit our project page at: https://umautobots.github.io/lihi_gs

Jingyu Song, Haoyu Ma, Onur Bagoren et al.

Underwater simulators offer support for building robust underwater perception solutions. Significant work has recently been done to develop new simulators and to advance the performance of existing underwater simulators. Still, there remains room for improvement on physics-based underwater sensor modeling and rendering efficiency. In this paper, we propose OceanSim, a high-fidelity GPU-accelerated underwater simulator to address this research gap. We propose advanced physics-based rendering techniques to reduce the sim-to-real gap for underwater image simulation. We develop OceanSim to fully leverage the computing advantages of GPUs and achieve real-time imaging sonar rendering and fast synthetic data generation. We evaluate the capabilities and realism of OceanSim using real-world data to provide qualitative and quantitative results. The code and detailed documentation are made available on the project website to support the marine robotics community: https://umfieldrobotics.github.io/OceanSim.

Pou-Chun Kung, Skanda Harisha, Ram Vasudevan et al.

High-Fidelity 3D scene reconstruction plays a crucial role in autonomous driving by enabling novel data generation from existing datasets. This allows simulating safety-critical scenarios and augmenting training datasets without incurring further data collection costs. While recent advances in radiance fields have demonstrated promising results in 3D reconstruction and sensor data synthesis using cameras and LiDAR, their potential for radar remains largely unexplored. Radar is crucial for autonomous driving due to its robustness in adverse weather conditions like rain, fog, and snow, where optical sensors often struggle. Although the state-of-the-art radar-based neural representation shows promise for 3D driving scene reconstruction, it performs poorly in scenarios with significant radar noise, including receiver saturation and multipath reflection. Moreover, it is limited to synthesizing preprocessed, noise-excluded radar images, failing to address realistic radar data synthesis. To address these limitations, this paper proposes RadarSplat, which integrates Gaussian Splatting with novel radar noise modeling to enable realistic radar data synthesis and enhanced 3D reconstruction. Compared to the state-of-the-art, RadarSplat achieves superior radar image synthesis (+3.4 PSNR / 2.6x SSIM) and improved geometric reconstruction (-40% RMSE / 1.5x Accuracy), demonstrating its effectiveness in generating high-fidelity radar data and scene reconstruction. A project page is available at https://umautobots.github.io/radarsplat.

Spencer Carmichael, Rahul Agrawal, Ram Vasudevan et al.

Recognizing places from an opposing viewpoint during a return trip is a common experience for human drivers. However, the analogous robotics capability, visual place recognition (VPR) with limited field of view cameras under 180 degree rotations, has proven to be challenging to achieve. To address this problem, this paper presents Same Place Opposing Trajectory (SPOT), a technique for opposing viewpoint VPR that relies exclusively on structure estimated through stereo visual odometry (VO). The method extends recent advances in lidar descriptors and utilizes a novel double (similar and opposing) distance matrix sequence matching method. We evaluate SPOT on a publicly available dataset with 6.7-7.6 km routes driven in similar and opposing directions under various lighting conditions. The proposed algorithm demonstrates remarkable improvement over the state-of-the-art, achieving up to 91.7% recall at 100% precision in opposing viewpoint cases, while requiring less storage than all baselines tested and running faster than all but one. Moreover, the proposed method assumes no a priori knowledge of whether the viewpoint is similar or opposing, and also demonstrates competitive performance in similar viewpoint cases.

Advaith V. Sethuraman, Max Rucker, Onur Bagoren et al.

In this paper, we present SonarSplat, a novel Gaussian splatting framework for imaging sonar that demonstrates realistic novel view synthesis and models acoustic streaking phenomena. Our method represents the scene as a set of 3D Gaussians with acoustic reflectance and saturation properties. We develop a novel method to efficiently rasterize Gaussians to produce a range/azimuth image that is faithful to the acoustic image formation model of imaging sonar. In particular, we develop a novel approach to model azimuth streaking in a Gaussian splatting framework. We evaluate SonarSplat using real-world datasets of sonar images collected from an underwater robotic platform in a controlled test tank and in a real-world river environment. Compared to the state-of-the-art, SonarSplat offers improved image synthesis capabilities (+3.2 dB PSNR) and more accurate 3D reconstruction (77% lower Chamfer Distance). We also demonstrate that SonarSplat can be leveraged for azimuth streak removal.

Anja Sheppard, Katherine A. Skinner

In this work, we propose the use of Ground Penetrating Radar (GPR) for rover localization on Mars. Precise pose estimation is an important task for mobile robots exploring planetary surfaces, as they operate in GPS-denied environments. Although visual odometry provides accurate localization, it is computationally expensive and can fail in dim or high-contrast lighting. Wheel encoders can also provide odometry estimation, but are prone to slipping on the sandy terrain encountered on Mars. Although traditionally a scientific surveying sensor, GPR has been used on Earth for terrain classification and localization through subsurface feature matching. The Perseverance rover and the upcoming ExoMars rover have GPR sensors already equipped to aid in the search of water and mineral resources. We propose to leverage GPR to aid in Mars rover localization. Specifically, we develop a novel GPR-based deep learning model that predicts 1D relative pose translation. We fuse our GPR pose prediction method with inertial and wheel encoder data in a filtering framework to output rover localization. We perform experiments in a Mars analog environment and demonstrate that our GPR-based displacement predictions both outperform wheel encoders and improve multi-modal filtering estimates in high-slip environments. Lastly, we present the first dataset aimed at GPR-based localization in Mars analog environments, which will be made publicly available at https://umfieldrobotics.github.io/marslgpr.

Parker Ewen, Hao Chen, Seth Isaacson et al.

This paper introduces a novel approach to uncertainty quantification for radiance fields by leveraging higher-order moments of the rendering equation. Uncertainty quantification is crucial for downstream tasks including view planning and scene understanding, where safety and robustness are paramount. However, the high dimensionality and complexity of radiance fields pose significant challenges for uncertainty quantification, limiting the use of these uncertainty quantification methods in high-speed decision-making. We demonstrate that the probabilistic nature of the rendering process enables efficient and differentiable computation of higher-order moments for radiance field outputs, including color, depth, and semantic predictions. Our method outperforms existing radiance field uncertainty estimation techniques while offering a more direct, computationally efficient, and differentiable formulation without the need for post-processing. Beyond uncertainty quantification, we also illustrate the utility of our approach in downstream applications such as next-best-view (NBV) selection and active ray sampling for neural radiance field training. Extensive experiments on synthetic and real-world scenes confirm the efficacy of our approach, which achieves state-of-the-art performance while maintaining simplicity.

Onur Bagoren, Marc Micatka, Katherine A. Skinner et al.

When navigating and interacting in challenging environments where sensory information is imperfect and incomplete, robots must make decisions that account for these shortcomings. We propose a novel method for quantifying and representing such perceptual uncertainty in 3D reconstruction through occupancy uncertainty estimation. We develop a framework to incorporate it into grasp selection for autonomous manipulation in underwater environments. Instead of treating each measurement equally when deciding which location to grasp from, we present a framework that propagates uncertainty inherent in the multi-view reconstruction process into the grasp selection. We evaluate our method with both simulated and the real world data, showing that by accounting for uncertainty, the grasp selection becomes robust against partial and noisy measurements. Code will be made available at https://onurbagoren.github.io/PUGS/

Jingyu Song, Zhenxin Li, Shiyi Lan et al.

Benchmarking autonomous driving planners to align with human judgment remains a critical challenge, as state-of-the-art metrics like the Extended Predictive Driver Model Score (EPDMS) lack context awareness in nuanced scenarios. To address this, we introduce DriveCritic, a novel framework featuring two key contributions: the DriveCritic dataset, a curated collection of challenging scenarios where context is critical for correct judgment and annotated with pairwise human preferences, and the DriveCritic model, a Vision-Language Model (VLM) based evaluator. Fine-tuned using a two-stage supervised and reinforcement learning pipeline, the DriveCritic model learns to adjudicate between trajectory pairs by integrating visual and symbolic context. Experiments show DriveCritic significantly outperforms existing metrics and baselines in matching human preferences and demonstrates strong context awareness. Overall, our work provides a more reliable, human-aligned foundation to evaluating autonomous driving systems.

Advaith V. Sethuraman, Onur Bagoren, Harikrishnan Seetharaman et al.

Mobile robots operating indoors must be prepared to navigate challenging scenes that contain transparent surfaces. This paper proposes a novel method for the fusion of acoustic and visual sensing modalities through implicit neural representations to enable dense reconstruction of transparent surfaces in indoor scenes. We propose a novel model that leverages generative latent optimization to learn an implicit representation of indoor scenes consisting of transparent surfaces. We demonstrate that we can query the implicit representation to enable volumetric rendering in image space or 3D geometry reconstruction (point clouds or mesh) with transparent surface prediction. We evaluate our method's effectiveness qualitatively and quantitatively on a new dataset collected using a custom, low-cost sensing platform featuring RGB-D cameras and ultrasonic sensors. Our method exhibits significant improvement over state-of-the-art for transparent surface reconstruction.

Spencer Carmichael, Austin Buchan, Mani Ramanagopal et al.

Conventional cameras employed in autonomous vehicle (AV) systems support many perception tasks, but are challenged by low-light or high dynamic range scenes, adverse weather, and fast motion. Novel sensors, such as event and thermal cameras, offer capabilities with the potential to address these scenarios, but they remain to be fully exploited. This paper introduces the Novel Sensors for Autonomous Vehicle Perception (NSAVP) dataset to facilitate future research on this topic. The dataset was captured with a platform including stereo event, thermal, monochrome, and RGB cameras as well as a high precision navigation system providing ground truth poses. The data was collected by repeatedly driving two ~8 km routes and includes varied lighting conditions and opposing viewpoint perspectives. We provide benchmarking experiments on the task of place recognition to demonstrate challenges and opportunities for novel sensors to enhance critical AV perception tasks. To our knowledge, the NSAVP dataset is the first to include stereo thermal cameras together with stereo event and monochrome cameras. The dataset and supporting software suite is available at: https://umautobots.github.io/nsavp

Kento Tomita, Katherine A. Skinner, Koki Ho

Hazard detection is critical for enabling autonomous landing on planetary surfaces. Current state-of-the-art methods leverage traditional computer vision approaches to automate the identification of safe terrain from input digital elevation models (DEMs). However, performance for these methods can degrade for input DEMs with increased sensor noise. In the last decade, deep learning techniques have been developed for various applications. Nevertheless, their applicability to safety-critical space missions has often been limited due to concerns regarding their outputs' reliability. In response to these limitations, this paper proposes an application of the Bayesian deep-learning segmentation method for hazard detection. The developed approach enables reliable, safe landing site detection by: (i) generating simultaneously a safety prediction map and its uncertainty map via Bayesian deep learning and semantic segmentation; and (ii) using the uncertainty map to filter out the uncertain pixels in the prediction map so that the safe site identification is performed only based on the certain pixels (i.e., pixels for which the model is certain about its safety prediction). Experiments are presented with simulated data based on a Mars HiRISE digital terrain model by varying uncertainty threshold and noise levels to demonstrate the performance of the proposed approach.

Abigail R. Azari, John B. Biersteker, Ryan M. Dewey et al.

Machine learning (ML) methods can expand our ability to construct, and draw insight from large datasets. Despite the increasing volume of planetary observations, our field has seen few applications of ML in comparison to other sciences. To support these methods, we propose ten recommendations for bolstering a data-rich future in planetary science.

Alexandra Carlson, Katherine A. Skinner, Ram Vasudevan et al.

Performance on benchmark datasets has drastically improved with advances in deep learning. Still, cross-dataset generalization performance remains relatively low due to the domain shift that can occur between two different datasets. This domain shift is especially exaggerated between synthetic and real datasets. Significant research has been done to reduce this gap, specifically via modeling variation in the spatial layout of a scene, such as occlusions, and scene environmental factors, such as time of day and weather effects. However, few works have addressed modeling the variation in the sensor domain as a means of reducing the synthetic to real domain gap. The camera or sensor used to capture a dataset introduces artifacts into the image data that are unique to the sensor model, suggesting that sensor effects may also contribute to domain shift. To address this, we propose a learned augmentation network composed of physically-based augmentation functions. Our proposed augmentation pipeline transfers specific effects of the sensor model -- chromatic aberration, blur, exposure, noise, and color temperature -- from a real dataset to a synthetic dataset. We provide experiments that demonstrate that augmenting synthetic training datasets with the proposed learned augmentation framework reduces the domain gap between synthetic and real domains for object detection in urban driving scenes.

Junming Zhang, Katherine A. Skinner, Ram Vasudevan et al.

Recent work has shown that convolutional neural networks (CNNs) can be applied successfully in disparity estimation, but these methods still suffer from errors in regions of low-texture, occlusions and reflections. Concurrently, deep learning for semantic segmentation has shown great progress in recent years. In this paper, we design a CNN architecture that combines these two tasks to improve the quality and accuracy of disparity estimation with the help of semantic segmentation. Specifically, we propose a network structure in which these two tasks are highly coupled. One key novelty of this approach is the two-stage refinement process. Initial disparity estimates are refined with an embedding learned from the semantic segmentation branch of the network. The proposed model is trained using an unsupervised approach, in which images from one half of the stereo pair are warped and compared against images from the other camera. Another key advantage of the proposed approach is that a single network is capable of outputting disparity estimates and semantic labels. These outputs are of great use in autonomous vehicle operation; with real-time constraints being key, such performance improvements increase the viability of driving applications. Experiments on KITTI and Cityscapes datasets show that our model can achieve state-of-the-art results and that leveraging embedding learned from semantic segmentation improves the performance of disparity estimation.

Alexandra Carlson, Katherine A. Skinner, Ram Vasudevan et al.

Recent work has focused on generating synthetic imagery to increase the size and variability of training data for learning visual tasks in urban scenes. This includes increasing the occurrence of occlusions or varying environmental and weather effects. However, few have addressed modeling variation in the sensor domain. Sensor effects can degrade real images, limiting generalizability of network performance on visual tasks trained on synthetic data and tested in real environments. This paper proposes an efficient, automatic, physically-based augmentation pipeline to vary sensor effects --chromatic aberration, blur, exposure, noise, and color cast-- for synthetic imagery. In particular, this paper illustrates that augmenting synthetic training datasets with the proposed pipeline reduces the domain gap between synthetic and real domains for the task of object detection in urban driving scenes.

Jie Li, Katherine A. Skinner, Ryan M. Eustice et al.

This paper reports on WaterGAN, a generative adversarial network (GAN) for generating realistic underwater images from in-air image and depth pairings in an unsupervised pipeline used for color correction of monocular underwater images. Cameras onboard autonomous and remotely operated vehicles can capture high resolution images to map the seafloor, however, underwater image formation is subject to the complex process of light propagation through the water column. The raw images retrieved are characteristically different than images taken in air due to effects such as absorption and scattering, which cause attenuation of light at different rates for different wavelengths. While this physical process is well described theoretically, the model depends on many parameters intrinsic to the water column as well as the objects in the scene. These factors make recovery of these parameters difficult without simplifying assumptions or field calibration, hence, restoration of underwater images is a non-trivial problem. Deep learning has demonstrated great success in modeling complex nonlinear systems but requires a large amount of training data, which is difficult to compile in deep sea environments. Using WaterGAN, we generate a large training dataset of paired imagery, both raw underwater and true color in-air, as well as depth data. This data serves as input to a novel end-to-end network for color correction of monocular underwater images. Due to the depth-dependent water column effects inherent to underwater environments, we show that our end-to-end network implicitly learns a coarse depth estimate of the underwater scene from monocular underwater images. Our proposed pipeline is validated with testing on real data collected from both a pure water tank and from underwater surveys in field testing. Source code is made publicly available with sample datasets and pretrained models.