

Do vision models still recognise an object once its usual neighbours are gone? A dataset and study of context reliance.


PhD student in computer vision & machine learning · MPI for Informatics & ISTA
I am an ELLIS PhD student at the Max Planck Institute for Informatics and ISTA, advised by Bernt Schiele and Francesco Locatello. My research centres on representation learning for vision and vision-language foundation models. I am especially interested in how these models can be used for visual understanding tasks such as detection, segmentation and localisation, and in making deep networks interpretable.
Before the PhD I was a Research Fellow at Microsoft Research India, building low-cost smartphone diagnostics for eye care, and a research intern at Adobe. I did my MS and B.Tech at IIIT Hyderabad with P. J. Narayanan.
When not working on my research, I like to play the piano 🎹 and guitar 🎸, listen to music 🎧, read non-fiction 📚, drive motorcycles 🏍️, and go for a run 🏃 or hike 🥾. I am also really fascinated by paradoxes ♾️ (you can find some here), and I wish I had Hermione's Time-Turner ⏳ to do much more in a day. See more at Personal.


Do vision models still recognise an object once its usual neighbours are gone? A dataset and study of context reliance.




One label-free alignment step turns DINOv3 or SigLIP2 into an inherently interpretable B-cos model.


How the final layer is trained (<10% of parameters) shapes post-hoc explanations; simple fixes improve them markedly.
@unpublished{gairola2026disparq,
author = {Adam Pardyl and Siddhartha Gairola and Sukrut Rao and Adam Wr{\'o}bel and Bartosz Zieli{\'n}ski and Bernt Schiele and Dawid Rymarczyk},
title = {{DisParQ}: Self-Supervised Part Concepts for Interpretable Vision Foundation Models},
note = {Under submission},
year = {2026}
}
@inproceedings{gairola2026stranger,
author = {Siddhartha Gairola and Jiahao Xie and Anna Kukleva and Francesco Locatello and Bernt Schiele},
title = {Stranger Things: When Objects Appear Without Their Typical Neighbours},
booktitle = {Conference on Neural Information Processing Systems (NeurIPS)},
year = {2026}
}
@inproceedings{gairola2026dave,
author = {Adam Wr{\'o}bel and Siddhartha Gairola and Jacek Tabor and Bernt Schiele and Bartosz Zieli{\'n}ski and Dawid Rymarczyk},
title = {{DAVE}: Distribution-aware Attribution via {ViT} Gradient Decomposition},
booktitle = {International Conference on Machine Learning (ICML)},
year = {2026},
url = {https://arxiv.org/abs/2602.06613}
}
Spotlight at the HOW Workshop and poster at the XAI4CV Workshop, CVPR 2026.
@inproceedings{gairola2026aloe,
author = {Raphael Maser and Siddhartha Gairola and Sukrut Rao and Bernt Schiele},
title = {Align Once to Explain: Feature Alignment for Scalable B-cosification of Foundational Vision Transformers},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2026},
url = {https://cvpr.thecvf.com/virtual/2026/poster/40400}
}
@inproceedings{gairola2025probe,
author = {Siddhartha Gairola and Moritz B{\"o}hle and Francesco Locatello and Bernt Schiele},
title = {How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations},
booktitle = {International Conference on Learning Representations (ICLR)},
year = {2025},
url = {https://openreview.net/pdf?id=57NfyYxh5f}
}
@inproceedings{ganatra2025smartkcpp,
author = {Vaibhav Ganatra and Siddhartha Gairola and Pallavi Joshi and Anand Balasubramaniam and Kaushik Murali and Anusha Varadharajan and Bhavana Mallikarjuna and Nipun Kwatra and Mohit Jain},
title = {{SmartKC}++: Improving Performance of Smartphone-Based Corneal Topographers},
booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
year = {2025},
url = {https://openaccess.thecvf.com/content/WACV2025/html/Ganatra_SmartKC_Improving_Performance_of_Smartphone-Based_Corneal_Topographers_WACV_2025_paper.html}
}
@inproceedings{gairola2022keratoconus,
author = {Siddhartha Gairola and Pallavi Joshi and Anand Balasubramaniam and Kaushik Murali and Nipun Kwatra and Mohit Jain},
title = {Keratoconus Classifier for Smartphone-based Corneal Topographer},
booktitle = {IEEE Engineering in Medicine and Biology Conference (EMBC)},
year = {2022},
url = {https://arxiv.org/abs/2205.03702}
}
@article{aggarwal2022retinoscopy,
author = {Aditya Aggarwal and Siddhartha Gairola and Uddeshya Upadhyay and Akshay P. Vasishta and Diwakar Rao and Aditya Goyal and Kaushik Murali and Nipun Kwatra and Mohit Jain},
title = {Towards Automating Retinoscopy for Refractive Error Diagnosis},
journal = {Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT)},
volume = {6},
number = {3},
year = {2022},
url = {https://arxiv.org/abs/2208.05552}
}
@article{gairola2021smartkc,
author = {Siddhartha Gairola and Murtuza Bohra and Nadeem Shaheer and Navya Jayaprakash and Pallavi Joshi and Anand Balasubramaniam and Kaushik Murali and Nipun Kwatra and Mohit Jain},
title = {{SmartKC}: Smartphone-based Corneal Topographer for Keratoconus Detection},
journal = {Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT)},
volume = {5},
number = {4},
year = {2021},
url = {https://sidgairo18.github.io/smartkc.pdf}
}
@inproceedings{gairola2021respirenet,
author = {Siddhartha Gairola and Francis Tom and Nipun Kwatra and Mohit Jain},
title = {{RespireNet}: A Deep Neural Network for Accurately Detecting Abnormal Lung Sounds in Limited Data Setting},
booktitle = {IEEE Engineering in Medicine and Biology Conference (EMBC)},
year = {2021},
url = {https://arxiv.org/pdf/2011.00196.pdf}
}
@inproceedings{gairola2020simpropnet,
author = {Siddhartha Gairola and Ayush Chopra and Mayur Hemani and {Balaji K.}},
title = {{SimPropNet}: Improved Similarity Propagation for Few-Shot Image Segmentation},
booktitle = {International Joint Conference on Artificial Intelligence (IJCAI)},
year = {2020},
url = {https://www.ijcai.org/Proceedings/2020/0080.pdf}
}
@inproceedings{gairola2020style,
author = {Siddhartha Gairola and Rajvi Shah and P. J. Narayanan},
title = {Unsupervised Image Style Embeddings for Retrieval and Recognition Tasks},
booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
year = {2020},
url = {https://openaccess.thecvf.com/content_WACV_2020/papers/Gairola_Unsupervised_Image_Style_Embeddings_for_Retrieval_and_Recognition_Tasks_WACV_2020_paper.pdf}
}
@inproceedings{kumar2018clickbait,
author = {Vaibhav Kumar and Dhruv Khattar and Siddhartha Gairola and Yash Kumar Lal and Vasudeva Varma},
title = {Identifying Clickbait: A Multi-Strategy Approach Using Neural Networks},
booktitle = {ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)},
year = {2018},
url = {https://arxiv.org/pdf/1710.01507.pdf}
}
@inproceedings{rawat2018sky,
author = {Saumya Rawat and Siddhartha Gairola and Rajvi Shah and P. J. Narayanan},
title = {Find Me a Sky: A Data-Driven Method for Color-Consistent Sky Search and Replacement},
booktitle = {International Conference on Multimedia Modeling (MMM)},
year = {2018},
url = {https://sidgairo18.github.io/sky.pdf}
}
*equal contribution · click bibtex on an entry to copy or download its citation
ICML 2023–26NeurIPS 2022–23, 2025ICLR 2022–24CVPR 2024–26ECCV 2024ICCV 2025IHCI 2021