Abstract: A noise filtering device, method, and chip for a dynamic vision sensor (DVS), and an event-driven vision device, and an electronic device are provided. In order to meet multi-dimensional requirements of a denoising solution such as power consumption, delay, storage resources, calculation resources, denoising effect, applicable range of noise types, and hardware implementation difficulty, in this disclosure, global space-domain and time-domain informations are simultaneously aggregated in a shift register, or latest time-domain informations are collected through a space domain, and finally, whether an event generated by a DVS is a noise event is judged on the basis of clustering of space-domain informations and the time-domain informations. While achieving excellent denoising performance, this disclosure has outstanding advantages in the aforementioned multiple dimensions because of advantages of parallel and simple calculation and no dependence on a DVS resolution, etc.

Abstract: A frame image conversion spike system is disclosed. In order to solve the technical problem of high power consumption when converting a frame image into a spike event, a frame rate or difference frame frequency is adaptively adjusted according to the amount of effective pixel change between frame images, so as to reduce system power consumption and adaptively respond to scene application needs. In order to solve the issue of low inference accuracy after converting frame images into spike events and directly applying them to an SNN processor, a global randomization scheme is proposed considering the randomization process of images as a whole. It has opened up all the links of applying a frame image sensor to the SNN processor and provided an alternative to an event camera. This is suitable for a brain-inspired chip and AIoT field.