C3Net: A cross-modal collaborative calibration of features for object detection using frames and events | Synapse