Search for: All records

Coastal farmlands in the eastern United States of America (USA) are increasingly suffering from rising soil salinity, rendering them unsuitable for economically productive agriculture. Saltwater intrusion (SWI) into the groundwater reservoir or soil salinization can result in land cover modification (e.g. reduced plant growth) or land cover conversion. Two primary examples of such land cover conversion are farmland to marsh or farmland to salt patches with no vegetation growth. However, due to varying spatial granularity of these conversions, it is challenging to quantify these land covers over a large geographic scale. To address this challenge, we evaluated a non-linear spectral unmixing approach with a Random Forest (RF) algorithm to quantify fractional abundance of salt patch and marshes. Using Sentinel-2 imagery from 2022, we generated gridded datasets for salt patches and marshes across the Delmarva Peninsula, and the associated uncertainty. Moreover, we developed two new spectral indices to enhance the spectral unmixing accuracy: the Normalized Difference Salt Patch Index (NDSPI) and the Modified Salt Patch Index (MSPI). We constructed two sets of ten RF models: one for salt patches and the other for marshes, achieving high (>99 %) training and testing accuracies for classification. The consistently high accuracy and low error values across different model runs demonstrate the method's reliability for classifying spectrally similar land cover classes in the mid-Atlantic region and beyond. Validation metrics for sub-pixel fractional abundances in the salt model revealed a moderate R-squared value of 0.50, and a high R-squared value of 0.90 for the marsh model. Our method complements labor-intensive field-based salinity measurements by offering a reproducible method that can be repeated annually and scaled up to cover large geographic regions. 

In this work, we propose a multi-objective decision making framework that accommodates different user preferences over objectives, where preferences are learned via policy comparisons. Our model consists of a known Markov decision process with a vector-valued reward function, with each user having an unknown preference vector that expresses the relative importance of each objective. The goal is to efficiently compute a near-optimal policy for a given user. We consider two user feedback models. We first address the case where a user is provided with two policies and returns their preferred policy as feedback. We then move to a different user feedback model, where a user is instead provided with two small weighted sets of representative trajectories and selects the preferred one. In both cases, we suggest an algorithm that finds a nearly optimal policy for the user using a number of comparison queries that scales quasilinearly in the number of objectives. 

Saltwater intrusion on coastal farmlands can render productive land unsuitable for agricultural activities. While the visible extent of salt-impacted land provides a useful saltwater intrusion proxy, it is challenging to identify in early stages. Moreover, associated ecological and economic impacts are often underestimated as reduced crop yields in farmlands surrounding salt patches are difficult to quantify. Here we develop a high-resolution (1 m) dataset showing salt patches on farm fringes and quantify the extent of salt-impacted lands across the Delmarva Peninsula, United States. Our method is transferable to other regions across and beyond the mid-Atlantic with similar saltwater intrusion issues, such as Georgia and the Carolinas. Our results show that between 2011 and 2017, visible salt patches almost doubled and 8,096 ha of farmlands converted to marsh—another saltwater intrusion consequence. Field-based electrical conductivity measurements show elevated salinity values hundreds of metres from visible salt patches, indicating the broader extent of at-risk farmlands. More farmland areas were within 200 m of a visible salt patch in 2017 compared to 2011, a rise ranging between 68% in Delaware and 93% in Maryland. On the basis of assumed 100% profit loss in at-risk farmlands within a 200 m buffer around salt patches in 2016–2017, the range of economic losses was estimated between US$39.4 million and US$107.5 million annually, under 100% soy or corn counterfactuals, respectively. 

A robot’s deployment environment often involves perceptual changes that differ from what it has experienced during training. Standard practices such as data augmentation attempt to bridge this gap by augmenting source images in an effort to extend the support of the training distribution to better cover what the agent might experience at test time. In many cases, however, it is impossible to know test-time distribution- shift a priori, making these schemes infeasible. In this paper, we introduce a general approach, called Invariance through Latent Alignment (ILA), that improves the test-time performance of a visuomotor control policy in deployment environments with unknown perceptual variations. ILA performs unsupervised adaptation at deployment-time by matching the distribution of latent features on the target domain to the agent’s prior experience, without relying on paired data. Although simple, we show that this idea leads to surprising improvements on a variety of challenging adaptation scenarios, including changes in lighting conditions, the content in the scene, and camera poses. We present results on calibrated control benchmarks in simulation—the distractor control suite—and a physical robot under a sim-to-real setup. Video and code available at: https: //invariance-through-latent-alignment.github.io 

Camera calibration is integral to robotics and computer vision algorithms that seek to infer geometric properties of the scene from visual input streams. In practice, calibration is a laborious procedure requiring specialized data collection and careful tuning. This process must be repeated whenever the parameters of the camera change, which can be a frequent occurrence for mobile robots and autonomous vehicles. In contrast, self-supervised depth and ego-motion estimation approaches can bypass explicit calibration by in-ferring per-frame projection models that optimize a view-synthesis objective. In this paper, we extend this approach to explicitly calibrate a wide range of cameras from raw videos in the wild. We propose a learning algorithm to regress per-sequence calibration parameters using an efficient family of general camera models. Our procedure achieves self-calibration results with sub-pixel reprojection error, outperforming other learning-based methods. We validate our approach on a wide variety of camera geometries, including perspective, fisheye, and catadioptric. Finally, we show that our approach leads to improvements in the downstream task of depth estimation, achieving state-of-the-art results on the EuRoC dataset with greater computational efficiency than contemporary methods. 

Contemporary approaches to perception, planning, estimation, and control have allowed robots to operate robustly as our remote surrogates in uncertain, unstructured environments. This progress now creates an opportunity for robots to operate not only in isolation, but also with and alongside humans in our complex environments. Realizing this opportunity requires an efficient and flexible medium through which humans can communicate with collaborative robots. Natural language provides one such medium, and through significant progress in statistical methods for natural-language understanding, robots are now able to interpret a diverse array of free-form navigation, manipulation, and mobile-manipulation commands. However, most contemporary approaches require a detailed, prior spatial-semantic map of the robot’s environment that models the space of possible referents of an utterance. Consequently, these methods fail when robots are deployed in new, previously unknown, or partially-observed environments, particularly when mental models of the environment differ between the human operator and the robot. This paper provides a comprehensive description of a novel learning framework that allows field and service robots to interpret and correctly execute natural-language instructions in a priori unknown, unstructured environments. Integral to our approach is its use of language as a “sensor”—inferring spatial, topological, and semantic information implicit in natural-language utterances and then exploiting this information to learn a distribution over a latent environment model. We incorporate this distribution in a probabilistic, language grounding model and infer a distribution over a symbolic representation of the robot’s action space, consistent with the utterance. We use imitation learning to identify a belief-space policy that reasons over the environment and behavior distributions. We evaluate our framework through a variety of different navigation and mobile-manipulation experiments involving an unmanned ground vehicle, a robotic wheelchair, and a mobile manipulator, demonstrating that the algorithm can follow natural-language instructions without prior knowledge of the environment. 

Shared autonomy provides an effective framework for human-robot collaboration that takes advantage of the complementary strengths of humans and robots to achieve common goals. Many existing approaches to shared autonomy make restrictive assumptions that the goal space, environment dynamics, or human policy are known a priori, or are limited to discrete action spaces, preventing those methods from scaling to complicated real world environments. We propose a model-free, residual policy learning algorithm for shared autonomy that alleviates the need for these assumptions. Our agents are trained to minimally adjust the human’s actions such that a set of goal-agnostic constraints are satisfied. We test our method in two continuous control environments: Lunar Lander, a 2D flight control domain, and a 6-DOF quadrotor reaching task. In experiments with human and surrogate pilots, our method significantly improves task performance without any knowledge of the human’s goal beyond the constraints. These results highlight the ability of model-free deep reinforcement learning to realize assistive agents suited to continuous control settings with little knowledge of user intent. 

We introduce a novel approach to endowing neural networks with emergent, long-term, large-scale memory. Distinct from strategies that connect neural networks to external memory banks via intricately crafted controllers and hand-designed attentional mechanisms, our memory is internal, distributed, co-located alongside computation, and implicitly addressed, while being drastically simpler than prior efforts. Architecting networks with multigrid structure and connectivity, while distributing memory cells alongside computation throughout this topology, we observe the emergence of coherent memory subsystems. Our hierarchical spatial organization, parameterized convolutionally, permits efficient instantiation of large-capacity memories, while multigrid topology provides short internal routing pathways, allowing convolutional networks to efficiently approximate the behavior of fully connected networks. Such networks have an implicit capacity for internal attention; augmented with memory, they learn to read and write specific memory locations in a dynamic data-dependent manner. We demonstrate these capabilities on exploration and mapping tasks, where our network is able to self-organize and retain long-term memory for trajectories of thousands of time steps. On tasks decoupled from any notion of spatial geometry: sorting, associative recall, and question answering, our design functions as a truly generic memory and yields excellent results.