Programmable black phosphor image sensor for broadband optoelectronic edge computing

Multifunctional infrared image sensor based on a programmable black phosphorus phototransistor (bP-PPT) array.

Abstract

“Image sensors with internal computing capability enable in-sensor computing that can significantly reduce communication latency and power consumption for machine vision in distributed systems and robotics. Two-dimensional semiconductors have many advantages in the realization of such intelligent vision sensors due to their tunable electrical and optical properties and their possibility of heterogeneous integration. Here, we report a multifunctional infrared image sensor based on a programmable black phosphorus phototransistor (bP-PPT) array. By electrically and optically controlling the charges stored in the gate dielectric layers, the electrical conductance and photoreactivity of bP-PPT can be programmed locally or remotely with 5-bit precision to implement a convolutional neural network (CNN) in the sensor. The sensor array can receive optical images transmitted over a wide infrared spectral range and perform inference computation to process and recognize the images with 92% accuracy. The demonstrated bP image sensor network can be scaled up to build a more complex vision sensory neural network, which will find many promising applications for distributed and remote multispectral sensing.

Find Open Access spec sheet here. Published 03-2022.

Lee, S., Peng, R., Wu, C. et al. Programmable black phosphor image sensor for broadband optoelectronic edge computing. Common Nat 13, 1485 (2022). https://doi.org/10.1038/s41467-022-29171-1

Michael C. Garrison