mlsanityData to 7 Oct 2026

Deep autoregressive modeling for land use land cover

arXiv 2401.01395 ↗First posted 2 Jan 2024cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Land use / land cover (LULC) modeling is a challenging task due to long-range dependencies between geographic features and distinct spatial patterns related to topography, ecology, and human development. We identify a close connection between modeling of spatial patterns of land use and the task of image inpainting from computer vision and conduct a study of a modified PixelCNN architecture with approximately 19 million parameters for modeling LULC. In comparison with a benchmark spatial statistical model, we find that the former is capable of capturing much richer spatial correlation patterns such as roads and water bodies but does not produce a calibrated predictive distribution, suggesting the need for additional tuning. We find evidence of predictive underdispersion with regard to important ecologically-relevant land use statistics such as patch count and adjacency which can be ameliorated to some extent by manipulating sampling variability.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 53 references that are in the map. First posted on arXiv 2 Jan 2024.

Builds on

23 of 53 references.

References · 23

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2017On Calibration of Modern Neural Networkscited by 4,860
  4. 2018Neural Ordinary Differential Equationscited by 3,733
  5. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,609
  6. 2016WaveNet: A Generative Model for Raw Audiocited by 1,710
  7. 2019Generating Diverse High-Fidelity Images with VQ-VAE-2cited by 1,376
  8. 2016Pixel Recurrent Neural Networkscited by 814
  9. 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 635
  10. 2017PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modificationscited by 415
  11. 2022Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioningcited by 362
  12. 2022Diffusion models as plug-and-play priorscited by 230
11 more
  1. 2019Pluralistic Image Completioncited by 162
  2. 2019Discrete Flows: Invertible Generative Models of Discrete Datacited by 68
  3. 2016GANS for Sequences of Discrete Elements with the Gumbel-softmax Distributioncited by 47
  4. 2021Hex2vec -- Context-Aware Embedding H3 Hexagons with OpenStreetMap Tagscited by 22
  5. 2019Hierarchical Autoregressive Image Models with Auxiliary Decoderscited by 17
  6. 2018Generating Realistic Geology Conditioned on Physical Measurements with Generative Adversarial Networkscited by 12
  7. 2022Resolving label uncertainty with implicit posterior modelscited by 8
  8. 2020Graphical Gaussian Process Models for Highly Multivariate Spatial Datacited by 6
  9. 2020Generative Autoregressive Ensembles for Satellite Imagery Manipulation Detectioncited by 6
  10. 2018Probabilistic Semantic Inpainting with Pixel Constrained CNNscited by 2
  11. 2022Large-Scale Auto-Regressive Modeling Of Street Networkscited by 2

Built on by

Not cited yet.