Skip to main content
. 2022 Feb 20;143:105298. doi: 10.1016/j.compbiomed.2022.105298

Table 4.

The inference timings on four different specification android devices for trained tensorflow model converted to flatbuffer.

Device Inference Timing (4 Threads) Inference Timing (6 Threads) Inference Timing (8 Threads)
Nokia 5.1 Plus 291 ms 273 ms 276 ms
Xiaomi Note4 303 ms 270 ms 270 ms
Samsung A 30 276 ms 253 ms 259 ms
Samsung Galaxy S2 312 ms 281 ms 282 ms