Research

1) A computational account of topography in the occipitotemporal cortex via domain-general pressures:

Collaborators: Talia Konkle

• Doshi, F. R., & Konkle, T. (2023). Cortical topographic motifs emerge in a self-organized map of object space. Science Advances, 9(25), eade8187.

• Doshi, F. R., & Konkle, T. (2023). Face-deprived networks show distributed but not clustered face-selective maps. Journal of Vision, 23(9), 5435-5435.

• Doshi, F., & Konkle, T. (2021). Organizational motifs of cortical responses to objects emerge in topographic projections of deep neural networks. Journal of Vision, 21(9), 2226-2226.
  Link to talk presented at the Vision Sciences Society 2021 Conference

2) Mechanisms of Contour Integration in Humans and Machines:

Collaborators: Talia Konkle and George Alvarez

• Doshi, F. R., Konkle, T., & Alvarez, G.A. (2024). A feedforward mechanism for human-like contour integration. Paper In Prep, 2024.

• Doshi, F. R., Konkle, T., & Alvarez, G.A. (2023). Feedforward Neural Networks can capture Human-like Perceptual and Behavioral Signatures of Contour Integration. In Cognitive Computational Neuroscience (CCN), 2023.

• Doshi, F., Konkle, T., & Alvarez, G.A. (2022). Human-like signatures of contour integration in deep neural networks.
  Link to talk presented at the Vision Sciences Society 2022 Conference

3) Mid-level Shape Perception in Deep Neural Networks

• Doshi, F. R., Konkle, T., & Alvarez, G.A. (2024). Configural-Shape Representation in Deep Neural Networks. Paper In Cognitive Computational Neuroscience (CCN), 2024.

• Doshi, F. R., Konkle, T., & Alvarez, G.A. (2024). Quantifying the Quality of Shape and Texture in Deep Neural Network Models. In Vision Science Society, 2024

4) Perceptual features as optimal proxies for intuitive physical reasoning:

Collaborators: Colin Conwell and George Alvarez

• Conwell, C., Doshi, F., Alvarez, G.A.(2019). Shared Representations of Stability in Humans, Supervised, & Unsupervised Neural Networks. In Shared Visual Representations in Human and Machine Intelligence. SVRHM workshop at NeurIPS 2019.

• Conwell, C., Doshi, F., Alvarez, G.A.(2019). Human-Like Judgments of Stability Emerge from Purely Perceptual Features: Evidence from Supervised and Unsupervised Deep Neural Networks. In Proceedings of the 3rd Conference on Cognitive Computational Neuroscience (CCN), 2019.

5) Capacity limits in visual working memory:

Collaborators: Hrag Pailian and George Alvarez

• Doshi, F., Pailian, H., & Alvarez, G. A. (2020). Using Deep Convolutional Neural Networks to Examine the Role of Representational Similarity in Visual Working Memory. Journal of Vision, 20(11), 149-149.
  Link to poster walkthrough presented at Vision Sciences Society 2020 Conference

Also check out Hrag’s talk here!

6) Previous work:

• Liu, Yc., Wu, H., Mayeshiba, T. et al. (2022). Machine learning predictions of irradiation embrittlement in reactor pressure vessel steels. In NPJ Computational Materials, 2022.

• Kanakasabapathy, M., Thirumalaraju, P., Kandula, H., Doshi, F., Sivakumar, A., Kartik, D., Gupta, R., Pooniwala, R., Branda, J., Tsibris, A., Kuritzkes, D., Petrozza, J., Bormann, C., Shafiee H. (2021). Adaptive adversarial neural networks for the analysis of lossy and domainshifted datasets of medical images. In Nature Biomedical Engineering, 2021.

• Thirumalaraju, P., Bormann, CL., Kanakasabapathy, M., Doshi, F., Souter, I., Dimitriadis, I., Shafiee, H.(2018). Automated sperm morpshology testing using artificial intelligence. In Fertility and sterility, 2018