Citizen Science

The current era of astronomy is defined by an unprecedented explosion of data, driven by wide-field surveys, space missions, and time-domain facilities. While advances in artificial intelligence and machine learning have significantly improved our ability to analyse large datasets, many scientific problems, particularly those involving complex morphologies and rare features, still benefit greatly from human visual inspection and intuition.

Citizen science provides a powerful framework to combine human pattern recognition with algorithmic approaches, enabling efficient exploration of large astronomical datasets while simultaneously engaging the public in frontline scientific research.

The “One Million Galaxies (OMG)” Project

I led the astronomy citizen-science initiative “One Million Galaxies (OMG)”, launched under the Pune Knowledge Cluster (PKC). The project is designed to identify and classify morphological features in galaxies using deep imaging data, particularly from wide-field surveys such as Subaru.

Scientific Objectives:

• Identification of galaxy structures such as bars, rings, lenses, tidal features, and shells

• Characterisation of faint and low-surface-brightness features associated with galaxy interactions and mergers

• Creation of a statistically robust, human-verified training set for machine-learning applications

• Integration of citizen classifications with automated pipelines for large survey analysis

Online Introductory Lectures

1. One Million Galaxies Launch: National Science Day 28 Feb 2022 -- Play Video

2. Finding features in Galaxies: Training session slot1 -- Play Video

3. Finding features in Galaxies: Training session slot2 -- Play Video

4. Finding features in Galaxies: Training session slot3 -- Play Video

5. Finding features in Galaxies:  Training session slot4 -- Play Video

6. One Million Galaxies: Enhanced Layout -- Play Video

Pilot Phase and Outcomes

The pilot phase of the OMG project demonstrated the feasibility and scientific value of large-scale citizen participation:

• ~100 participants from diverse backgrounds

• ~15,000 galaxies visually inspected

• Successful identification of multiple morphological features difficult to detect with automated methods alone

The results validated the role of citizen science as a complementary tool to AI, particularly for tasks where training data are sparse or features are subtle.

Human–AI Synergy

A core philosophy of the OMG project is the synergistic integration of citizen science and machine learning. Human classifications serve multiple critical roles:

• Providing ground truth labels for training and validating deep-learning models

• Highlighting rare or unexpected features that automated pipelines may miss

• Enabling iterative refinement of algorithms through expert–citizen feedback loops

This hybrid approach enhances both scientific reliability and algorithmic performance, and is central to my broader research programme in data-driven astronomy.

Education and Capacity Building

Beyond its scientific goals, the citizen-science programme is designed as an educational and capacity-building platform. Participants include:

• Undergraduate and postgraduate students

• Amateur astronomers and astronomy clubs

• Educators and science enthusiasts

The project exposes participants to real astronomical data and research workflows, fostering scientific literacy, data skills, and critical thinking.

Integration with Research and Teaching

Citizen science activities are closely integrated with my research, student supervision, and teaching. Students at different levels participate in project development, data analysis, and interpretation, gaining hands-on experience with:

• Large astronomical datasets

• Visual classification techniques

• AI-assisted analysis workflows

This integration ensures that citizen science contributes directly to both scientific output and training.

Broader Impact and Future Vision

The long-term vision of the OMG project is to scale the effort to one million galaxies, creating one of the largest human-classified morphological datasets. Such a dataset will be invaluable for:

• Statistical studies of galaxy evolution

• Training next-generation AI models

• Preparing for data volumes from upcoming large surveys

More broadly, I view citizen science as an essential component of inclusive and sustainable astronomy, enabling meaningful public participation while accelerating scientific discovery.