Skip to Main Content
Table 3: 

Retrieval latency for one query with an image collection of 50k or 1M images (with pre-encoded images) using a single GPU/ CPU. Batch size for cross-encoding of the query with the images is 512. CPU is an Intel Xeon Gold 6154.

ModelNVIDIA V100CPU
50k1M50k1M
BE 16ms 37ms 0.2s 1.6s 
Sep/Joint+Coop 74ms 94ms 6s 13s 
CE 2min 36min 2.4h 47h 
ModelNVIDIA V100CPU
50k1M50k1M
BE 16ms 37ms 0.2s 1.6s 
Sep/Joint+Coop 74ms 94ms 6s 13s 
CE 2min 36min 2.4h 47h 
Close Modal

or Create an Account

Close Modal
Close Modal