| Abstract: | Radio observations are essential for understanding galaxy formation and evolution, yet low-frequency interferometric observations are often hindered by radio frequency interference (RFI) and system failures, making data processing a time-intensive challenge. With next-generation radio telescopes producing increasingly large datasets, the demand for automated data processing solutions has grown critical. We present GARUDA (Generic AI-based GMRT-tUned Radio Data Analysis pipeline), a novel automated pipeline designed for uGMRT data reduction. Written in Python and utilizing modular CASA for calibration, GARUDA includes GNET, our custom Deep Learning based RFI detection model. With only two tunable parameters, GNET ensures flexibility and ease of use across diverse observations and frequency bands. The pipeline handles system issues and performs RFI excision, producing high quality calibrated data ready for imaging. GARUDA processes 10-12 GB GSB data in 20-30 minutes and approximately 400 GB GWB data in under three hours on standard GPU workstations, achieving rapid and reliable results. In this talk, I will discuss GARUDA's capabilities and showcase results, including some of the deepest GMRT radio continuum images at the L-band, HI emission in galaxies, and one of the most sensitive galactic HI absorption lines (using frequency switching observation with GWB).
|
