Samir Chowdhury bio photo

Samir Chowdhury

Senior computer vision engineer at Bear Flag Robotics

Twitter Google Scholar Github

Research

Preprints

  1. Hypergraph Co-Optimal Transport: Metric and Categorical Properties. arXiv.
    With Tom Needham, Ethan Semrad, Bei Wang, and Youjia Zhou.

Publications

2022

  1. Temporal Mapper: transition networks in simulated and real neural dynamics. Accepted to Network Neuroscience.
    With Mengsen Zhang and Manish Saggar.
  2. Distances and isomorphism between networks: stability and convergence of network invariants. Journal of Applied and Computational Topology.
    With Facundo Mémoli.
  3. NeuMapper: A scalable computational framework for multiscale exploration of the brain’s dynamical organization. Network Neuroscience.
    With Caleb Geniesse and Manish Saggar.
  4. Path homologies of motifs and temporal network representations. Applied Network Science.
    With Steve Huntsman and Matvey Yutin.

    2021

  5. Quantized Gromov-Wasserstein. ECML.
    With David Miller and Tom Needham.
  6. Generalized Spectral Clustering via Gromov-Wasserstein Learning. AISTATS. Supplementary material.
    With Tom Needham.

    2020

  7. Path homology and temporal networks. Complex Networks 2020.
    With Steve Huntsman and Matvey Yutin.
  8. Gromov-Wasserstein Averaging in a Riemannian Framework. CVPR workshops (DiffCVML).
    With Tom Needham.
  9. New families of stable simplicial filtration functors. Topology and its Applications.
    With Nate Clause, Facundo Mémoli, Jose Ángel Sánchez, and Zoe Wellner.

    2019

  10. Path homologies of deep feedforward networks. IEEE ICMLA.
    With Tom Gebhart, Steve Huntsman, and Matvey Yutin.
  11. The Gromov-Wasserstein distance between networks and stable network invariants. Information and Inference.
    With Facundo Mémoli.
  12. Metric and Topological Approaches to Network Data Analysis. PhD Thesis.

    2018

  13. A functorial Dowker theorem and persistent homology of asymmetric networks. Journal of Applied and Computational Topology.
    With Facundo Mémoli.
  14. The importance of forgetting: Limiting memory improves recovery of topological characteristics from neural data. Plos One.
    With Bowen Dai and Facundo Mémoli.
  15. Explicit geodesics in Gromov-Hausdorff space. Electronic Research Announcements.
    With Facundo Mémoli.
  16. Persistent path homology of directed networks. SODA.
    With Facundo Mémoli.

    2016

  17. Hierarchical representations of network data with optimal distortion bounds. IEEE Asilomar.
    With Facundo Mémoli and Zane Smith.
  18. Persistent homology of directed networks. IEEE Asilomar.
    With Facundo Mémoli.
  19. Improved error bounds for tree representations of metric spaces. NeurIPS.
    With Facundo Mémoli and Zane Smith.
  20. Distances between directed networks and applications. IEEE ICASSP.
    With Facundo Mémoli.

    2015

  21. Metric structures on networks and applications. IEEE Allerton.
    With Facundo Mémoli.

Older preprints