US 12,335,518 B1
Optimizing video data upload for dashcam event reviews
Jing Wang, Toronto (CA); Narendran Rajan, San Diego, CA (US); William Nathan John Hurst, Seattle, WA (US); William Evan Welbourne, Seattle, WA (US); Greg Allen Cummings, San Francisco, CA (US); and Siddharth Heroor, Peoria, AZ (US)
Assigned to Samsara Inc., San Francisco, CA (US)
Filed by Samsara Inc., San Francisco, CA (US)
Filed on Dec. 9, 2022, as Appl. No. 18/064,224.
Int. Cl. H04N 19/59 (2014.01); G06V 20/40 (2022.01); H04N 19/167 (2014.01)
CPC H04N 19/59 (2014.11) [G06V 20/44 (2022.01); H04N 19/167 (2014.11)] 20 Claims
OG exemplary drawing
 
1. A method, performed by an IoT device, comprising:
capturing an original resolution video segment from a camera of the IoT device;
detecting an event in the original resolution video segment;
in response to the detecting the event:
storing the original resolution video segment in a memory of the IoT device,
converting the original resolution video segment to a low-resolution video segment by downscaling the original resolution video segment, and
sending the low-resolution video segment to a cloud backend, where the cloud backend is capable of making the low-resolution video segment available to a first reviewer for review, where the first reviewer flags the low-resolution video segment for further review;
receiving a request from the cloud backend for the original resolution video segment, where a second reviewer sends the request in response to the flagging by the first reviewer; and
sending the requested original resolution video segment to the cloud backend, where the cloud backend is capable of making the original resolution video segment available to the second reviewer for review.