Lunar Technosignatures

Overview

Uncovering anomalies on the lunar surface is crucial for understanding the Moon's geological and astronomical history. This study seeks to enhance anomaly detection on the Moon and replace the time-consuming manual data search process by testing anomaly detection methods using the Apollo landing sites. The landing sites are advantageous as they are both anomalous and precisely locatable, enabling a rigorous assessment of each approach.

Evaluated Architectures

AnoViT

Vision Transformer Autoencoder

The AnoViT method uses an Autoencoder approach to detect anomalies. It uses a Vision Transformer (ViT) as an encoder, and the resulting features are used to reconstruct the input image using a Convolutional Neural Network (CNN) decoder. The reconstruction error reveals anomalous regions that the model has never encountered during training.

ViT Encoder CNN Decoder Reconstruction Error Unsupervised

EfficientAD

Student-Teacher Framework

This approach uses a student-teacher framework along with an Autoencoder to identify anomalies at both local and global scales. A pre-trained teacher network provides feature targets; the student network learns to mimic these on normal data. During inference, significant differences between the student and teacher outputs are considered indicators of anomalies.

Student–Teacher Autoencoder Multi-scale ResNet18

Cut&Paste

Synthetic Anomaly Generation

This method converts the unsupervised anomaly detection problem into an artificially constructed supervised classification task. Synthetic anomalies are generated by cutting patches from normal images and pasting them back with augmentations. It was trained with five different transformations including mean patch, whiten blob, darken blob, and a darken path transform.

Synthetic Anomalies Supervised 5 Augmentations

Experimental Results

The results demonstrate that our approach outperforms previous publications in accurately predicting landing site artifacts and technosignatures at the Apollo 15 and 17 landing sites. The AnoViT model provides the most precise global anomaly map, showing a sharp and comprehensive representation of the lunar surface.

Qualitative Results

One important factor to consider in the classification accuracy is the number of anomalies detected around the landing sites (false positives). To make these false positive examples visible, we provide a global anomaly map in Figure 2 showing the whole testing area in one image, instead of only a single patch out of the whole test dataset, which is shown in Figure 1. In Figure 1, the local anomaly maps are clear and easy to understand. However, the complexity increases when looking at a global scale (Figure 2). It is worth noting that the descent stage is surrounded by more noise, especially in the EfficientAD and Cut&Paste models. On the other hand, the AnoViT model provides the most precise global anomaly map, showing a sharp and comprehensive representation of the lunar surface. This suggests that AnoViT might be a more effective tool for identifying technosignatures on the Moon as it creates a sharper reconstruction of the input image.

(a) Original

(b) Ground truth

(c) AnoViT

(d) Cut&Paste

(e) EffAD Local

(f) EffAD Global

Figure 1: Local anomaly maps from the descending stage of the Apollo 15 mission.

(a) EfficientAD Global

(b) EfficientAD Local

(c) Cut&Paste

(d) AnoViT

Figure 2: Global anomaly maps covering the complete test dataset of the Apollo 15 landing site.