We explore the potential of large-scale noisily labeled data to enhance feature learning by pretraining semantic segmentation models within a multimodal framework for geospatial applications. We propose a novel cross-modal sample selection (CromSS) method, a weakly supervised pretraining strategy designed to improve feature representations through cross-modal consistency and noise mitigation techniques. Unlike conventional pretraining approaches, CromSS exploits massive amounts of noisy and easy-to-come-by labels for improved feature learning beneficial to semantic segmentation tasks. We investigate middle and late fusion strategies to optimize the multimodal pretraining architecture design. We also introduce a cross-modal sample selection module to mitigate the adverse effects of label noise, which employs a cross-modal entangling strategy to refine the estimated confidence masks within each modality to guide the sampling process. Additionally, we introduce a spatial–temporal label smoothing technique to counteract overconfidence for enhanced robustness against noisy labels. To validate our approach, we assembled the multimodal dataset, NoLDO-S12, which consists of a large-scale noisy label subset from Google’s Dynamic World (DW) dataset for pretraining and two downstream subsets with high-quality labels from Google DW and OpenStreetMap (OSM) for transfer learning. Experimental results on two downstream tasks and the publicly available DFC2020 dataset demonstrate that when effectively utilized, the low-cost noisy labels can significantly enhance feature learning for segmentation tasks. The data, codes, and pretrained weights are freely available athttps://github.com/zhu-xlab/CromSS.

Compared to supervised deep learning, self-supervision provides remote sensing a tool to reduce the amount of exact, human-crafted geospatial annotations. While image-level information for unsupervised pretraining efficiently works for various classification downstream tasks, the performance on pixel-level semantic segmentation lags behind in terms of model accuracy. On the contrary, many easily available label sources (e.g., automatic labeling tools and land cover land use products) exist, which can provide a large amount of noisy labels for segmentation model training. In this work, we propose to exploit noisy semantic segmentation maps for model pretraining. Our experiments provide insights on robustness per network layer. The transfer learning settings test the cases when the pretrained encoders are fine-tuned for different label classes and decoders. The results from two datasets indicate the effectiveness of task-specific supervised pretraining with noisy labels. Our findings pave new avenues to improved model accuracy and novel pretraining strategies for efficient remote sensing image segmentation.

Local climate zones (LCZs) established a standard classification system to categorize the landscape universe for improved urban climate studies. Existing LCZ mapping is guided by human interaction with geographic information systems (GIS) or modelled from remote sensing (RS) data. GIS-based methods do not scale to large areas. However, RS-based methods leverage machine learning techniques to automatize LCZ classification from RS. Yet, RS-based methods require huge amounts of manual labels for training. We propose a novel LCZ mapping framework, termed AutoLCZ, to extract the LCZ classification parameters from high-resolution RS modalities. We study the definition of numerical rules designed to mimic the LCZ definitions. Those rules model geometric and surface cover parameters from LiDAR data. Correspondingly, we enable LCZ classification from RS data in a GIS-based scheme. The proposed AutoLCZ method has potential to reduce the human labor to acquire accurate metadata. At the same time, AutoLCZ sheds light on the physical interpretability of RS-based methods. In a proof-of-concept for New York City (NYC) we leverage airborne LiDAR surveys to model four LCZ parameters to distinguish eight LCZ types. The results indicate the potential of AutoLCZ as a promising avenue for large-scale LCZ mapping from RS data.

Preservation of the Nasca geoglyphs at the UNESCO World Heritage Site in Peru is urgent as natural and human impact accelerates. More frequent weather extremes such as flash-floods threaten Nasca geoglyphs. We demonstrate that runoff models based on (sub-)meter scale, LiDAR-derived digital elevation data can highlight AI-detected geoglyphs that are in danger of erosion. We recommend measures of mitigation to protect the famous “lizard”, “tree”, and “hand” geoglyphs located close by, or even cut by the Pan-American Highway.

Pretraining foundation models on large-scale satellite imagery has raised great interest in Earth observation. While most pretraining is conducted purely self-supervised, many land cover land use products that provide free and global annotations tend to be overlooked. To bridge this gap, we propose to exploit land-cover-generated multi-label annotations to guide supervised contrastive learning for Earth observation. We match the SSL4EO-S12 dataset with Dynamic World land cover maps and integrate image-level multi-label annotations. During pretraining, the label similarities between different images are calculated, and those with high similarity scores are pulled together in the embedding space. Experimental results on classification and segmentation downstream tasks demonstrate the effectiveness of the proposed method.

While the volume of remote sensing data is increasing daily, deep learning in Earth Observation faces lack of accurate annotations for supervised optimization. Crowdsourcing projects such as OpenStreetMap distribute the annotation load to their community. However, such annotation inevitably generates noise due to insufficient control of the label quality, lack of annotators, frequent changes of the Earth’s surface as a result of natural disasters and urban development, among many other factors. We presentAdaptively trIggered Online Object-wise correction (AIO2)to address annotation noise induced by incomplete label sets. AIO2 features anAdaptive Correction Trigger (ACT)module that avoids label correction when the model training under- or overfits, and anOnline Object-wise Correction (O2C)methodology that employs spatial information for automated label modification. AIO2 utilizes a mean teacher model to enhance training robustness with noisy labels to both stabilize the training accuracy curve for fitting in ACT and provide pseudo labels for correction in O2C. Moreover, O2C is implementedonlinewithout the need to store updated labels every training epoch. We validate our approach on two building footprint segmentation datasets with different spatial resolutions. Experimental results with varying degrees of building label noise demonstrate the robustness of AIO2. Source code will be available at https://github.com/zhu-xlab/AIO2.git.

Self-supervised pretraining on large-scale satellite data has raised great interest in building Earth observation (EO) foundation models. However, many important resources beyond pure satellite imagery, such as land-cover-land-use products that provide free global semantic information, as well as vision foundation models that hold strong knowledge of the natural world, are not widely studied. In this work, we show these free additional resources not only help resolve common contrastive learning bottlenecks but also significantly boost the efficiency and effectiveness of EO pretraining. Specifically, we first propose soft contrastive learning (SoftCon) that optimizes cross-scene soft similarity based on land-cover-generated multilabel supervision, naturally solving the issue of multiple positive samples and too strict positive matching in complex scenes. Second, we revisit and explore cross-domain continual pretraining for both multispectral and synthetic aperture radar (SAR) imagery, building efficient EO foundation models from strongest vision models such as DINOv2. Adapting simple weight-initialization and Siamese masking strategies into our SoftCon framework, we demonstrate impressive continual pretraining performance even when the input modalities are not aligned. Without prohibitive training, we produce multispectral and SAR foundation models that achieve significantly better results in 10 out of 11 downstream tasks than most existing SOTA models. For example, our ResNet50/ViT-S achieve 84.8/85.0 linear probing mAP scores on BigEarthNet-10%, which are better than most existing ViT-L models; under the same setting, our ViT-B sets a new record of 86.8 in multispectral, and 82.5 in SAR, the latter even better than many multispectral models. Dataset and models are available athttps://github.com/zhu-xlab/softcon.

Nature-based carbon sequestration solution have the potential to capture carbon dioxide from the atmosphere and store it in vegetation biomass or soil. Forests are covering around 30% of Earth’s land surface and combined with forest longevity, trees/soil have the potential to store carbon from decades to centuries. One key challenge is to develop methodologies for high-resolution measurements of carbon sequestered and assess year to year change. Here, we use deep neural network to generate a wall-to-wall map of AGB within the Continental USA (CONUS) with 30-meter spatial resolution for the year 2021. We combine radar and optical multispectral imagery, with a physical climate parameter of Solar Induced Fluorescence (SIF)-based Growth Primary Production (GPP). Validation results show that a masked variation of UNet has the lowest validation RMSE of 37.93 ± 1.36 Mg C/ha, as compared to 81.95 ± 0.01 Mg C/ha (linear regressor), 53.37 ± 0.05 Mg C/ha (gradient boosting), and 52.30 ± 0.03 Mg C/ha for random forest algorithm. Furthermore, models that learn from SIF-based GPP in addition to radar and optical imagery reduce validation RMSE by almost 10% and the standard deviation by 40%.

Hyperspectral image (HSI) classification is gaining a lot of momentum in present time because of high inherent spectral information within the images. However, these images suffer from the problem of curse of dimensionality and usually require a large number samples for tasks such as classification, especially in supervised setting. Recently, to effectively train the deep learning models with minimal labelled samples, the unlabeled samples are also being leveraged in self-supervised and semi-supervised setting. In this work, we leverage the idea of semi-supervised learning to assist the discriminative self-supervised pretraining of the models. The proposed method takes different augmented views of the unlabeled samples as input and assigns them the same pseudo-label corresponding to the labelled sample from the downstream task. We train our model on two HSI datasets, anemly Houston dataset (from data fusion contest, 2013) and Pavia university dataset, and show that the proposed approach performs better than self-supervised approach and supervised training.

Urban land use structures impact local climate conditions of metropolitan areas. To shed light on the mechanism of local climate wrt. urban land use, we present a novel, data-driven deep learning architecture and pipeline, DeepLCZChange, to correlate airborne LiDAR data statistics with the Landsat 8 satellite’s surface temperature product. A proof-of-concept numerical experiment utilizes corresponding remote sensing data for the city of New York to verify the cooling effect of urban forests.

Automatic label generation systems, which are capable to generate huge amounts of labels with limited human efforts, enjoy lots of potential in the deep learning era. These easy-to-come-by labels inevitably bear label noises due to a lack of human supervision and can bias model training to some inferior solutions. However, models can still learn some plausible features, before they start to overfit on noisy patterns. Inspired by this phenomenon, we propose a new Peaks fusion assisted EArly-Stopping (PEAS) approach for imagery segmentation with noisy labels, which is mainly composed of two parts. First, a fitting based early-stopping criterion is used to detect the turning phase from which models are about to mimic noise details. After that, a peaks fusion strategy is applied to select reliable models in the detection zone to generate final fusion results. Here, validation accuracies are utilized as indicators in model selection. The proposed method was evaluated on New York City dataset whose labels were automatically collected by a rule-based label generation system, thus noisy to some extent due to a lack of human supervision. The experimental results showed that the proposed PEAS method can achieve both promising statistical and visual results when trained with noisy labels.

Discovering ancient agricultural terraces in desert regions is important for the monitoring of long-term climate changes on the Earth’s surface. However, traditional ground surveys are both costly and limited in scale. With the increasing accessibility of aerial and satellite data, machine learning techniques bear large potential for the automatic detection and recognition of archaeological landscapes. In this paper, we propose a deep semantic model fusion method for ancient agricultural terrace detection. The input data includes aerial images and LiDAR generated terrain features in the Negev desert. Two deep semantic segmentation models, namely DeepLabv3+ and UNet, with EfficientNet backbone, are trained and fused to provide segmentation maps of ancient terraces and walls. The proposed method won the first prize in the International AI Archaeology Challenge. Codes are available at https://github.com/wangyi111/international-archaeologyai-challenge.

We present and evaluate a weakly-supervised methodology to quantify the spatiotemporal distribution of urban forests based on remotely sensed data with close-to-zero human interaction. Successfully training machine learning models for semantic segmentation typically depends on the availability of high-quality labels. We evaluate the benefit of high-resolution, three-dimensional point cloud data (LiDAR) as source of noisy labels in order to train models for the localization of trees in orthophotos. As proof of concept we sense Hurricane Sandy's impact on urban forests in Coney Island, New York City (NYC) and reference it to less impacted urban space in Brooklyn, NYC.

In order to estimate tree biomass, allometric equations take tree parameters such as tree height, wood density, circumference of trunk, and crown diameter as input parameters. Given that most of these quantities are challenging to be extracted from remote sensing data, we evaluate the option to approximate biomass by tree height only. We study our approach by evaluating linear regression, random forest, and Gaussian process regressor models when applied to the 2016 Jucker dataset. Results indicate that linear models fail to properly capture the relationship between biomass and tree height, but the Gaussian process regressor outperms the other two candidate models.

Self-supervised learning (SSL) has attracted much interest in remote sensing and Earth observation due to its ability to learn task-agnostic representations without human annotation. While most of the existing SSL works in remote sensing utilize ConvNet backbones and focus on a single modality, we explore the potential of vision transformers (ViTs) for joint SAR-optical representation learning. Based on DINO, a state-of-the-art SSL algorithm that distills knowledge from two augmented views of an input image, we combine SAR and optical imagery by concatenating all channels to a unified input. Subsequently, we randomly mask out channels of one modality as a data augmentation strategy. While training, the model gets fed optical-only, SAR-only, and SAR-optical image pairs learning both inner- and intra-modality representations. Experimental results employing the BigEarthNet-MM dataset demonstrate the benefits of both, the ViT backbones and the proposed multimodal SSL algorithm DINO-MM.

A key challenge of supervised learning is the availability of human-labeled data. We evaluate a big data processing pipeline to auto-generate labels for remote sensing data. It is based on rasterized statistical features extracted from surveys such as e.g. LiDAR measurements. Using simple combinations of the rasterized statistical layers, it is demonstrated that multiple classes can be generated at accuracies of ~ 0.9.As proof of concept, we utilize the big geo-data platform IBM PAIRS to dynamically generate such labels in dense urban areas with multiple land cover classes. The general method proposed here is platform independent, and it can be adapted to generate labels for other satellite modalities in order to enable machine learning on overhead imagery for land use classification and object detection.

Accurately and globally mapping human infrastructure is an important and challenging task with applications in routing, regulation compliance monitoring, and natural disaster response management etc.. In this paper we present progress in developing an algorithmic pipeline and distributed compute system that automates the process of map creation using high resolution aerial images. Unlike previous studies, most of which use datasets that are available only in a few cities across the world, we utilizes publicly available imagery and map data, both of which cover the contiguous United States (CONUS). We approach the technical challenge of inaccurate and incomplete training data adopting state-of-the-art convolutional neural network architectures such as the U-Net and the CycleGAN to incrementally generate maps with increasingly more accurate and more complete labels of man-made infrastructure such as roads and houses. Since scaling the mapping task to CONUS calls for parallelization, we then adopted an asynchronous distributed stochastic parallel gradient descent training scheme to distribute the computational workload onto a cluster of GPUs with nearly linear speed-up.

We present a remote sensing pipeline that processes LiDAR (Light Detection And Ranging) data through machine & deep learning for the application of archeological feature detection on big geo-spatial data platforms such as e.g. IBM PAIRS Geoscope [1], [2].Today, archeologists get overwhelmed by the task of visually surveying huge amounts of (raw) LiDAR data in order to identify areas of interest for inspection on the ground. We showcase a software system pipeline that results in significant savings in terms of expert productivity while missing only a small fraction of the artifacts.Our work employs artificial neural networks in conjunction with an efficient spatial segmentation procedure based on domain knowledge. Data processing is constraint by a limited amount of training labels and noisy LiDAR signals due to vegetation cover and decay of ancient structures. We aim at identifying geo-spatial areas with archeological artifacts in a supervised fashion allowing the domain expert to flexibly tune parameters based on her needs.

The assessment of the vegetation growth rate given remote sensing data is a challenging task in the Earth Observation sciences. LiDAR data acquisition is commonly used to extract height information at a given moment in time, however, the associated cost and complexity restrict continuous acquisitions. Frequently captured aerial imagery can be used to identify and separate vegetation from bare land, water, impervious surface, or built infrastructure. A combination of LiDAR data with aerial and radar imagery allows to track dynamic seasonal growth of vegetation around critical infrastructure such as power lines. We present a general framework that integrates tree identification and growth assessment around power lines with the goal to identify locations of high risk where trees potentially cause power outages.

We present an efficient clustering algorithm applicable to one-dimensional data such as e.g. a series of times-tamps. Given an expected frequency ΔT-1, we introduce an O(N)-efficient method of characterizing N events represented by an ordered series of timestamps t1, t2,..., tN. In practice, the method proves useful to e.g. identify time intervals of missing data or to locate isolated events. Moreover, we define measures to quantify a series of events by varying ΔT to e.g. determine the quality of an Internet of Things service.

IBM's Physical Analytics Integrated Data Repository and Services (PAIRS) is a geospatial Big Data service. PAIRS contains a massive amount of curated geospatial (or more precisely spatio-temporal) data from a large number of public and private data resources, and also supports user contributed data layers. PAIRS offers an easy-to-use platform for both rapid assembly and retrieval of geospatial datasets or performing complex analytics, lowering time-to-discovery significantly by reducing the data curation and management burden. In this paper, we review recent progress with PAIRS and showcase a few exemplary analytical applications which the authors are able to build with relative ease leveraging this technology.

Geospatial data volume exceeds hundreds of Petabytes and is increasing exponentially mainly driven by images/videos/data generated by mobile devices and high resolution imaging systems. Fast data discovery on historical archives and/or real time datasets is currently limited by various data formats that have different projections and spatial resolution, requiring extensive data processing before analytics can be carried out. A new platform called Physical Analytics Integrated Repository and Services (PAIRS) is presented that enables rapid data discovery by automatically updating, joining, and homogenizing data layers in space and time. Built on top of open source big data software, PAIRS manages automatic data download, data curation, and scalable storage while being simultaneously a computational platform for running physical and statistical models on the curated datasets. By addressing data curation before data being uploaded to the platform, multi-layer queries and filtering can be performed in real time. In addition, PAIRS offers a foundation for developing custom analytics. Towards that end we present two examples with models which are running operationally: (1) high resolution evapo-transpiration and vegetation monitoring for agriculture and (2) hyperlocal weather forecasting driven by machine learning for renewable energy forecasting.