Bio-mimetic high-speed target localization with fused frame and event vision for edge application

HIGHLIGHTS

SUMMARY

    Spiking neural_networks exhibit low power consumption in customized hardware platforms (Akopyan et_al, 2015; Davies et_al, 2018) by the exploitation of asynchronous decentralized tile-based designs. The state-of-the-art method for object detection uses convolutional neural_networks (CNN) due to its high accuracy (Zhao et_al, 2019; Jiao et_al, 2019). The latency can be reduced while preserving the accuracy by equipping more powerful computing hardware on the drones (Duisterhof et_al, 2019; Wyder et_al, 2019; Falanga et_al, 2020). Spiking neural_networks, therefore, lie in the region of low accuracy and low latency (Kim et_al, 2020; Cannici et_al . . .

     

    Logo ScioWire Beta black

    If you want to have access to all the content you need to log in!

    Thanks :)

    If you don't have an account, you can create one here.

     

Scroll to Top

Add A Knowledge Base Question !

+ = Verify Human or Spambot ?