Pulse sequence randomization method and device, brain-like chip and electronic equipment

Abstract

The invention relates to a pulse sequence randomization method and device, a brain-like chip and electronic equipment. In order to convert an image output by a frame image sensor or an environment signal collected by the sensor into a randomized target pulse sequence suitable for being processed by a pulse neural network processor, a pulse event address and a pulse event number are stored; and reading the corresponding pulse event address or the number of pulse events according to a random number generation module, thereby generating a target pulse sequence formed by the corresponding pulse events. The method has the technical effects of low hardware resource consumption, good randomization effect, contour enhancement and background noise suppression, and provides a new technical path for generating a pulse sequence with biological characteristics. The method is suitable for the field of brain-like chips.

Description

technical field [0001] The present invention relates to a pulse sequence randomization method, device, brain-inspired chip and electronic equipment, and in particular to a method, device, brain-inspired chip and electronic equipment for converting an image into a random pulse sequence. Background technique [0002] A dynamic vision sensor (Dynamic Vision Sensor, DVS) is a new type of sensor deployed with an array of independent light-sensitive pixels, which independently and asynchronously send pulse events to the next-level system (such as a brain-like chip) after the pixels sense the change in light intensity . Because of its ultra-high time resolution, it has attracted much attention, but there is still a certain gap between its signal-to-noise ratio, dark-light sensing capability, consistency, and measurability and actual application requirements. [0003] The traditional frame-based image sensor (such as CMOS sensor, CCD sensor, etc., referred to as frame image sensor)...

AI Technical Summary

Problems solved by technology

But its disadvantage is that the number of randomly exchanged or changed pulses must be limited, and it is difficult to guarantee the effect of randomization, especially when the time step is long; the cost of hardware implementation is relatively high; in addition, it does not mention the difference frame method to obtain the contour information, but only performs random pulse sequence conversion for each pixel value of an ordinary frame image, and the recognition accuracy of the pulse sequence obtained in this way is extremely low in the spike neural network

[0009] Existing technology 3 provides a region-of-interest extraction scheme based on difference frames, but does not provide a scheme for pulse sequence randomization

Images