CFHT UM2025 - Presentation Details
Abstract Details
Title: Revealing CLASSY TNOs with Deep Learning
Presenter: Preeti Cowan
Abstract:
We examine deep learning’s effectiveness for TNO detection in CLASSY (Classical and Large-A Solar SYstem) survey, a CFHT Large Program now in its third year. In recent years, progress in TNO population science has been driven by large-scale surveys, precipitating in an avalanche of data that has highlighted the need for new tools to successfully unpack the signal from the noise. Deep learning models can build useful representations directly from the data without a lot of hand-engineering, making them an ideal addition to a survey’s analysis pipeline. Here, we create composite images of the nightly CFHT MegaCam observations, leveraging the TNO sky motion in consecutive observations to extract a linear series of points (‘tracklets’). These tracklets form the basis of our training dataset of ~75,000 images. Several custom convolutional neural network-based models were trained to identify and locate these tracklets in the CLASSY data. The predictions from each of these networks were then combined to boost the predictive power and reduce the false negatives. This technique is effective for isolating the regions of interest in observations, reducing the load for computing the predicted orbital track and ultimately streamlining the detection pipeline.