More Like this

1. Partial Domain Adaptation Using Selective Representation Learning For Class-Weight Computation

   https://doi.org/10.1109/IEEECONF51394.2020.9443420  

   Choudhuri, Sandipan; Paul, Riti; Sen, Arunabha; Li, Baoxin; Venkateswara, Hemanth (January 2020, 2020 54th Asilomar Conference on Signals, Systems, and Computers)

   Matthews, MB (Ed.)

   The generalization power of deep-learning models is dependent on rich-labelled data. This supervision using large-scaled annotated information is restrictive in most realworld scenarios where data collection and their annotation involve huge cost. Various domain adaptation techniques exist in literature that bridge this distribution discrepancy. However, a majority of these models require the label sets of both the domains to be identical. To tackle a more practical and challenging scenario, we formulate the problem statement from a partial domain adaptation perspective, where the source label set is a super set of the target label set. Driven by the motivation that image styles are private to each domain, in this work, we develop a method that identifies outlier classes exclusively from image content information and train a label classifier exclusively on class-content from source images. Additionally, elimination of negative transfer of samples from classes private to the source domain is achieved by transforming the soft class-level weights into two clusters, 0 (outlier source classes) and 1 (shared classes) by maximizing the between-cluster variance between them. 

   more » « less
2. Addressing the problem of scale that emerges with habitat fragmentation

   https://doi.org/10.1111/geb.13658  

   Fletcher, Jr., Robert J.; Betts, Matthew G.; Damschen, Ellen I.; Hefley, Trevor J.; Hightower, Jessica; Smith, Thomas A. H.; Fortin, Marie‐Josée; Haddad, Nick M. (March 2023, Global Ecology and Biogeography)

   Abstract Fragmentation and scaleAlthough habitat loss has well‐known impacts on biodiversity, the effects of habitat fragmentation remain intensely debated. It is often argued that the effects of habitat fragmentation, or the breaking apart of habitat for a given habitat amount, can be understood only at the scale of entire landscapes composed of multiple habitat patches. Yet, fragmentation also impacts the size, isolation and habitat edge for individual patches within landscapes. Addressing the problem of scale on fragmentation effects is crucial for resolving how fragmentation impacts biodiversity. Scaling frameworkWe build upon scaling concepts in ecology to describe a framework that emphasizes three “dimensions” of scale in habitat fragmentation research: the scales of phenomena (or mechanisms), sampling and analysis. Using this framework, we identify ongoing challenges and provide guidance for advancing the science of fragmentation. ImplicationsWe show that patch‐ and landscape‐scale patterns arising from habitat fragmentation for a given amount of habitat are fundamentally related, leading to interdependencies among expected patterns arising from different scales of phenomena. Aggregation of information when increasing the grain of sampling (e.g., from patch to landscape) creates challenges owing to biases created from the modifiable areal unit problem. Consequently, we recommend that sampling strategies use the finest grain that captures potential underlying mechanisms (e.g., plot or patch). Study designs that can capture phenomena operating at multiple spatial extents offer the most promise for understanding the effects of fragmentation and its underlying mechanisms. By embracing the interrelationships among scales, we expect more rapid advances in our understanding of habitat fragmentation. 

   more » « less
3. Toward a large‐scale and deep phenological stage annotation of herbarium specimens: Case studies from temperate, tropical, and equatorial floras

   https://doi.org/10.1002/aps3.1233  

   Lorieul, Titouan; Pearson, Katelin D.; Ellwood, Elizabeth R.; Goëau, Hervé; Molino, Jean‐Francois; Sweeney, Patrick W.; Yost, Jennifer M.; Sachs, Joel; Mata‐Montero, Erick; Nelson, Gil; et al (March 2019, Applications in Plant Sciences)

   Premise of the StudyPhenological annotation models computed on large‐scale herbarium data sets were developed and tested in this study. MethodsHerbarium specimens represent a significant resource with which to study plant phenology. Nevertheless, phenological annotation of herbarium specimens is time‐consuming, requires substantial human investment, and is difficult to mobilize at large taxonomic scales. We created and evaluated new methods based on deep learning techniques to automate annotation of phenological stages and tested these methods on four herbarium data sets representing temperate, tropical, and equatorial American floras. ResultsDeep learning allowed correct detection of fertile material with an accuracy of 96.3%. Accuracy was slightly decreased for finer‐scale information (84.3% for flower and 80.5% for fruit detection). DiscussionThe method described has the potential to allow fine‐grained phenological annotation of herbarium specimens at large ecological scales. Deeper investigation regarding the taxonomic scalability of this approach is needed. 

   more » « less
4. Spectral Characteristics of the Dynamic World Land Cover Classification

   https://doi.org/10.3390/rs15030575  

   Small, Christopher; Sousa, Daniel (February 2023, Remote Sensing)

   The Dynamic World product is a discrete land cover classification of Sentinel 2 reflectance imagery that is global in extent, retrospective to 2015, and updated continuously in near real time. The classifier is trained on a stratified random sample of 20,000 hand-labeled 5 × 5 km Sentinel 2 tiles spanning 14 biomes globally. Since the training data are based on visual interpretation of image composites by both expert and non-expert annotators, without explicit spectral properties specified in the class definitions, the spectral characteristics of the classes are not obvious. The objective of this study is to quantify the physical distinctions among the land cover classes by characterizing the spectral properties of the range of reflectance present within each of the Dynamic World classes over a variety of landscapes. This is achieved by comparing both the eight-class probability feature space (excluding snow) and the maximum probability class assignment (label) distributions to continuous land cover fraction estimates derived from a globally standardized spectral mixture model. Standardized substrate, vegetation, and dark (SVD) endmembers are used to unmix nine Sentinel 2 reflectance tiles from nine spectral diversity hotspots for comparison between the SVD land cover fraction continua and the Dynamic World class probability continua and class assignments. The variance partition for the class probability feature spaces indicates that eight of these nine hotspots are effectively five-dimensional to 95% of variance. Class probability feature spaces of the hotspots all show a tetrahedral form with probability continua spanning multiple classes. Comparison of SVD land cover fraction distributions with maximum probability class assignments (labels) and probability feature space distributions reveal a clear distinction between (1) physically and spectrally heterogeneous biomes characterized by continuous gradations in vegetation density, substrate albedo, and structural shadow fractions, and (2) more homogeneous biomes characterized by closed canopy vegetation (forest) or negligible vegetation cover (e.g., desert, water). Due to the ubiquity of spectrally heterogeneous biomes worldwide, the class probability feature space adds considerable value to the Dynamic World maximum probability class labels by offering users the opportunity to depict inherently gradational heterogeneous landscapes otherwise not generally offered with other discrete thematic classifications. 

   more » « less
5. Multi-Class Imbalanced Graph Convolutional Network Learning

   https://doi.org/10.24963/ijcai.2020/398  

   Shi, Min; Tang, Yufei; Zhu, Xingquan; Wilson, David; Liu, Jianxun (July 2020, Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence (IJCAI))

   null (Ed.)

   Networked data often demonstrate the Pareto principle (i.e., 80/20 rule) with skewed class distributions, where most vertices belong to a few majority classes and minority classes only contain a handful of instances. When presented with imbalanced class distributions, existing graph embedding learning tends to bias to nodes from majority classes, leaving nodes from minority classes under-trained. In this paper, we propose Dual-Regularized Graph Convolutional Networks (DR-GCN) to handle multi-class imbalanced graphs, where two types of regularization are imposed to tackle class imbalanced representation learning. To ensure that all classes are equally represented, we propose a class-conditioned adversarial training process to facilitate the separation of labeled nodes. Meanwhile, to maintain training equilibrium (i.e., retaining quality of fit across all classes), we force unlabeled nodes to follow a similar latent distribution to the labeled nodes by minimizing their difference in the embedding space. Experiments on real-world imbalanced graphs demonstrate that DR-GCN outperforms the state-of-the-art methods in node classification, graph clustering, and visualization. 

   more » « less