We introduce a method to train vision-language models for remote-sensing images without using any textual annotations. Our key insight is to use co-located internet imagery taken on the ground as an intermediary for connecting remote-sensing images and language. Specifically, we train an image encoder for remote sensing images to align with the image encoder of CLIP using a large amount of paired internet and satellite images. Our unsupervised approach enables the training of a first-of-its-kind large-scale vision language model (VLM) for remote sensing images at two different resolutions. We show that these VLMs enable zero-shot, open-vocabulary image classification, retrieval, segmentation and visual question answering for satellite images. On each of these tasks, our VLM trained without textual annotations outperforms existing VLMs trained with supervision, with gains of up to 20% for classification and 80% for segmentation.
Geographic variance in satellite imagery impacts the ability of machine learning models to generalise to new regions. In this paper, we model geographic generalisation in medium resolution Landsat-8 satellite imagery as a continuous domain adaptation problem, demonstrating how models generalise better with appropriate domain knowledge. We develop a dataset spatially distributed across the entire continental United States, providing macroscopic insight into the effects of geography on crop classification in multi-spectral and temporally distributed satellite imagery. Our method demonstrates improved generalisability from 1) passing geographically correlated climate variables along with the satellite data to a Transformer model and 2) regressing on the model features to reconstruct these domain variables. Combined, we provide a novel perspective on geographic generalisation in satellite imagery and a simple-yet-effective approach to leverage domain knowledge. Code is available at: \url{https://github.com/samar-khanna/cropmap}
The process of capturing a well-composed photo is difficult and it takes years of experience to master. We propose a novel pipeline for an autonomous agent to automatically capture an aesthetic photograph by navigating within a local region in a scene. Instead of classical optimization over heuristics such as the rule-of-thirds, we adopt a data-driven aesthetics estimator to assess photo quality. A reinforcement learning framework is used to optimize the model with respect to the learned aesthetics metric. We train our model in simulation with indoor scenes, and we demonstrate that our system can capture aesthetic photos in both simulation and real world environments on a ground robot. To our knowledge, this is the first system that can automatically explore an environment to capture an aesthetic photo with respect to a learned aesthetic estimator.
Modern recognition systems require large amounts of supervision to achieve accuracy. Adapting to new domains requires significant data from experts, which is onerous and can become too expensive. Zero-shot learning requires an annotated set of attributes for a novel category. Annotating the full set of attributes for a novel category proves to be a tedious and expensive task in deployment. This is especially the case when the recognition domain is an expert domain. We introduce a new field-guide-inspired approach to zero-shot annotation where the learner model interactively asks for the most useful attributes that define a class. We evaluate our method on classification benchmarks with attribute annotations like CUB, SUN, and AWA2 and show that our model achieves the performance of a model with full annotations at the cost of a significantly fewer number of annotations. Since the time of experts is precious, decreasing annotation cost can be very valuable for real-world deployment.
We present PhySG, an end-to-end inverse rendering pipeline that includes a fully differentiable renderer and can reconstruct geometry, materials, and illumination from scratch from a set of RGB input images. Our framework represents specular BRDFs and environmental illumination using mixtures of spherical Gaussians, and represents geometry as a signed distance function parameterized as a Multi-Layer Perceptron. The use of spherical Gaussians allows us to efficiently solve for approximate light transport, and our method works on scenes with challenging non-Lambertian reflectance captured under natural, static illumination. We demonstrate, with both synthetic and real data, that our reconstructions not only enable rendering of novel viewpoints, but also physics-based appearance editing of materials and illumination.
We present a new framework for semantic segmentation without annotations via clustering. Off-the-shelf clustering methods are limited to curated, single-label, and object-centric images yet real-world data are dominantly uncurated, multi-label, and scene-centric. We extend clustering from images to pixels and assign separate cluster membership to different instances within each image. However, solely relying on pixel-wise feature similarity fails to learn high-level semantic concepts and overfits to low-level visual cues. We propose a method to incorporate geometric consistency as an inductive bias to learn invariance and equivariance for photometric and geometric variations. With our novel learning objective, our framework can learn high-level semantic concepts. Our method, PiCIE (Pixel-level feature Clustering using Invariance and Equivariance), is the first method capable of segmenting both things and stuff categories without any hyperparameter tuning or task-specific pre-processing. Our method largely outperforms existing baselines on COCO and Cityscapes with +17.5 Acc. and +4.5 mIoU. We show that PiCIE gives a better initialization for standard supervised training. The code is available at https://github.com/janghyuncho/PiCIE.
Reconstructing the shape and appearance of real-world objects using measured 2D images has been a long-standing problem in computer vision. In this paper, we introduce a new analysis-by-synthesis technique capable of producing high-quality reconstructions through robust coarse-to-fine optimization and physics-based differentiable rendering. Unlike most previous methods that handle geometry and reflectance largely separately, our method unifies the optimization of both by leveraging image gradients with respect to both object reflectance and geometry. To obtain physically accurate gradient estimates, we develop a new GPU-based Monte Carlo differentiable renderer leveraging recent advances in differentiable rendering theory to offer unbiased gradients while enjoying better performance than existing tools like PyTorch3D and redner. To further improve robustness, we utilize several shape and material priors as well as a coarse-to-fine optimization strategy to reconstruct geometry. We demonstrate that our technique can produce reconstructions with higher quality than previous methods such as COLMAP and Kinect Fusion.
The fashion sense -- meaning the clothing styles people wear -- in a geographical region can reveal information about that region. For example, it can reflect the kind of activities people do there, or the type of crowds that frequently visit the region (e.g., tourist hot spot, student neighborhood, business center). We propose a method to automatically create underground neighborhood maps of cities by analyzing how people dress. Using publicly available images from across a city, our method finds neighborhoods with a similar fashion sense and segments the map without supervision. For 37 cities worldwide, we show promising results in creating good underground maps, as evaluated using experiments with human judges and underground map benchmarks derived from non-image data. Our approach further allows detecting distinct neighborhoods (what is the most unique region of LA?) and answering analogy questions between cities (what is the "Downtown LA" of Bogota?).
A common strategy for improving model robustness is through data augmentations. Data augmentations encourage models to learn desired invariances, such as invariance to horizontal flipping or small changes in color. Recent work has shown that arbitrary style transfer can be used as a form of data augmentation to encourage invariance to textures by creating painting-like images from photographs. However, a stylized photograph is not quite the same as an artist-created painting. Artists depict perceptually meaningful cues in paintings so that humans can recognize salient components in scenes, an emphasis which is not enforced in style transfer. Therefore, we study how style transfer and paintings differ in their impact on model robustness. First, we investigate the role of paintings as style images for stylization-based data augmentation. We find that style transfer functions well even without paintings as style images. Second, we show that learning from paintings as a form of perceptual data augmentation can improve model robustness. Finally, we investigate the invariances learned from stylization and from paintings, and show that models learn different invariances from these differing forms of data. Our results provide insights into how stylization improves model robustness, and provide evidence that artist-created paintings can be a valuable source of data for model robustness.
Deep learning has paved the way for strong recognition systems which are often both trained on and applied to natural images. In this paper, we examine the give-and-take relationship between such visual recognition systems and the rich information available in the fine arts. First, we find that visual recognition systems designed for natural images can work surprisingly well on paintings. In particular, we find that interactive segmentation tools can be used to cleanly annotate polygonal segments within paintings, a task which is time consuming to undertake by hand. We also find that FasterRCNN, a model which has been designed for object recognition in natural scenes, can be quickly repurposed for detection of materials in paintings. Second, we show that learning from paintings can be beneficial for neural networks that are intended to be used on natural images. We find that training on paintings instead of natural images can improve the quality of learned features and we further find that a large number of paintings can be a valuable source of test data for evaluating domain adaptation algorithms. Our experiments are based on a novel large-scale annotated database of material depictions in paintings which we detail in a separate manuscript.