Publications

Acoustic Signal Processing

• On Adversarial Attacks In Acoustic Drone Localization– Tamir Shor, Chaim Baskin, Alex Bronstein

  • Paper
  • TL;DR – In this work we formulate adversarial attacks over agents performing acoustic localization, benchmark their effect over localization accuracy, and propose a novel adversarial defense algorithm adapted to this setting.

• Active propulsion noise shaping for multi-rotor aircraft localization –Gabriele Serussi, Tamir Shor, Tom Hirshberg, Chaim Baskin, Alex Bronstein (Proc. Int’l Conf. on Intelligent Robots and Systems (IROS) 2024, ICML Workshop on AI for Science 2024)

  • Paper
  • Code
  • TL;DR – In this work we develop a novel algorithm for the localization of MAVs (drones) using only the propulsion noise emitted by the rotors. We also propose active shaping of rotor phase modulation for better localization capability.

Medical Imaging

• T1-PILOT: Optimized Trajectories for T1 Mapping Acceleration – Tamir Shor, Moti Freiman, Chaim Baskin, Alex Bronstein

  • Paper
  • TL;DR – In this work we develop an algorithm for accelerated T1 Mapping, by levaraging learned, physically feasible, per-frame acquisition trajectories that are directly informed by the physical exponential decay model.

• Leveraging Latents for Efficient Thermography Classification and Segmentation – Tamir Shor, Chaim Baskin, Alex Bronstein (Proc. Medical Imaging with Deep Learning (MIDL), 2024)

  • Paper
  • Code
  • TL;DR – In this work we develop an efficient pipeline for the classification and segmentation of thermography data, and demonstrate the benefits of using a potent latent space over existing approaches.

• Multi PILOT: Learned Feasible Multiple Acquisition Trajectories for Dynamic MRI – Tamir Shor, Tomer Weiss, Dor Noti, Alex Bronstein (Proc. Medical Imaging with Deep Learning (MIDL), 2023)

  • Paper
  • Code
  • TL;DR – In this work we propose a novel algorithm for non-Cartesian deep MRI Compressed Sensing for MRI videos.

Computer Vision

• • Adversarial Robustness in Parameter-Space Classifiers –Tamir Shor, Ethan Fetaya, Chaim Baskin, Alex Bronstein (ICLR Workshop on Weight Space Learning, 2025)

    • Paper
    • Code
    • TL;DR – In this work explore adversarial robustness in classifiers operating directly in the weight-space of neural networks trained for implicit neural representation. We develop a suite of adversarial attacks adapted for this case, evaluate their performance over classification accuracy, and show evidence of inherent adversarial robustness apparent in parameter-space classifiers.

  • Single Image Test-Time Adaptation for Segmentation – Klara Janouskova, Tamir Shor, Chaim Baskin, Jiri Matas (Proc. Transaction on Machine Learning Research (TMLR), 2024)

    • Paper
    • Code
    • TL;DR – In this work we propose a novel method for Test-Time Adaptation in image segmentation, as well as benchmark existing approaches.