Learn from the brain, but shall not be limited to. The research of artificial intelligence has been inferenced dramatically by neuroscience, such as the representation of visual stimuli [1, 2] and visual attention [3, 4]. On the other hand, given that one of the ultimate goals of artificial intelligence is to exceed the human ability [5], modern computer algorithms do not necessarily have to mimic the brain [6]. Thus, our research philosophy is to incorporate biological knowledge into a computer algorithm whenever suitable, but not be bound to it.

Latest Research

One example of our bio-inspired machine learning is the study of the Sequential Probability Ratio Test, or SPRT [7]. SPRT is originally invented by Abraham Wald, later re-discovered as the algorithm that best explains the primate parietal lobe neurons’ activity [8]. The parietal neurons are thought to be engaged in evidence-accumulating processes in complex decision-making. We extended Wald’s algorithm to a more general case, using deep neural networks aided by density ratio estimation algorithms [9]. Coined as SPRT-TANDEM, our algorithm showed a competitive speed and accuracy under various real-world scenarios. Interested readers are invited to read the introductory article on GitHub and/or the original paper [10].

Research Interests

• Bio-inspired machine learning
• Sequential Probability Ratio Test (SPRT)
• Density ratio estimation
• Face
  • Face recognition (computer vision)
  • Face recognition (visual neuroscience)
  • Face presentation attack detection (PAD)
    
    

News

[Apr. 1st, 2024]　　

• Akinori F. Ebihara has been promoted to Principal Researcher.

[Feb. 16th, 2023]　　

• Our paper, “Toward Asymptotic Optimality: Sequential Unsupervised Regression of Density Ratio for Early Classification” was accepted to ICASSP 2023.

[Apr. 27th, 2022]　　

• Our paper, “Convolutional Neural Networks for Time-dependent Classification of Variable-length Time Series” was accepted to WCCI 2021 (IJCNN track) as an oral presentation.　　

[Jun. 2nd, 2021]　　

• Our paper, “Joint Feature Distribution Alignment Learning for NIR-VIS and VIS-VIS Face Recognition” was accepted to IJCB 2021.　　

[May 8th, 2021]　　

• Our paper, “Sequential Density Ratio Matrix Estimation for Speed-Accuracy Optimization” was accepted to ICML 2021.　　

[May 7th, 2021]

• Nikkei (日本経済新聞) and Nikkei XTech featured articles about our algorithm, SPRT-TANDEM.
• NEC released an official press release about the SPRT-TANDEM.

[Apr. 1st, 2021]

• Akinori F. Ebihara has been promoted to “Special Researcher.”
  *A few distinguished researchers that are selected to enter a performance-based reward system instead of the traditional seniority wage system.

[Jan. 13th, 2021]

• Our paper, “Sequential Density Ratio Estimation for Simultaneous Optimization of Speed and Accuracy” was accepted as a Spotlight presentation (top 5.57%) at ICLR 2021.

[Oct. 1st, 2020]

• Our paper, “Specular- and Diffuse-reflection-based Face Spoofing Detection for Mobile Devices” won IJCB 2020 Google Best Paper Award.

References

1. Hubel, D. H., Wiesel, T. N. RECEPTIVE FIELDS AND FUNCTIONAL ARCHITECTURE IN TWO NONSTRIATE VISUAL AREAS (18 AND 19) OF THE CAT. Journal of Neurophysiology 1965, 28 (2), 229–289.
2. Fukushima, K. Neocognitron: A Self-Organizing Neural Network Model for a Mechanism of Pattern Recognition Unaffected by Shift in Position. Biol. Cybernetics 1980, 36 (4), 193–202.
3. Posner, M. I., Petersen, S. E. The Attention System of the Human Brain. Annu. Rev. Neurosci. 1990, 13 (1), 25–42.
4. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., Polosukhin, I. Attention Is All You Need. 11.
5. Silver, D., Schrittwieser, J., Simonyan, K., Antonoglou, I., Huang, A., Guez, A., Hubert, T., Baker, L., Lai, M., Bolton, A., Chen, Y., Lillicrap, T., Hui, F., Sifre, L., van den Driessche, G., Graepel, T., Hassabis, D. Mastering the Game of Go without Human Knowledge. Nature 2017, 550 (7676), 354–359.
6. Nonaka, S., Majima, K., Aoki, S. C., Kamitani, Y. Brain Hierarchy Score: Which Deep Neural Networks Are Hierarchically Brain-Like? bioRxiv 2020.
7. Wald, A. Sequential Analysis. Wiley, Hoboken, NJ, USA, 1947.
8. Kira, S., Yang, T., Shadlen, M. N. A Neural Implementation of Wald’s Sequential Probability Ratio Test. Neuron 2015, 85 (4), 861–873.
9. Sugiyama, M., Suzuki, T., Kanamori, T. Density ratio estimation in machine learning. Cambridge University Press, 2012.
10. Ebihara, A.F., Miyagawa, T., Sakurai, K., Imaoka, H. Sequential Density Ratio Estimation for Simultaneous Optimization of Speed and Accuracy. ICLR 2021.