AI Based Object Redaction/Performance: Difference between revisions

From RidgeRun Developer Wiki
Line 313: Line 313:
{| class="wikitable" style="text-align:center;"
{| class="wikitable" style="text-align:center;"
|-
|-
! colspan="19" | Image 1080p- 7 faces
! colspan="25" | Image 1080p- 7 faces
|-
|-
| Platform
| Platform
| colspan="18" | Xavier AGX(8-cores 30W Mode )
| colspan="24" | Xavier AGX(8-cores 30W Mode )
|-
|-
| Method
| Method
| colspan="6" style="background-color:#ffce93;" | Convert
| colspan="6" style="background-color:#ffce93;" | Convert
| colspan="6" style="background-color:#9aff99;" | Detect
| colspan="6" style="background-color:#9aff99;" | Detect
| colspan="6" style="background-color:#abe9ff;" | Track
| colspan="6" style="background-color:#96fffb;" | Track
| colspan="6" style="background-color:#ffffc7;" | Redact
| colspan="6" style="background-color:#fffc9e;" | Redact
|-
|-
| Num threads
| Num threads
Line 329: Line 329:
| colspan="3" style="background-color:#9aff99;" | 1
| colspan="3" style="background-color:#9aff99;" | 1
| colspan="3" style="background-color:#9aff99;" | 4
| colspan="3" style="background-color:#9aff99;" | 4
| colspan="3" style="background-color:#abe9ff;" | 1
| colspan="3" style="background-color:#96fffb;" | 1
| colspan="3" style="background-color:#abe9ff;" | 4
| colspan="3" style="background-color:#96fffb;" | 4
| colspan="3" style="background-color:#ffffc7;" | 1
| colspan="3" style="background-color:#fffc9e;" | 1
| colspan="3" style="background-color:#ffffc7;" | 4
| colspan="3" style="background-color:#fffc9e;" | 4
|-
|-
| Stat
| Stat
Line 347: Line 347:
| style="background-color:#9aff99;" | Processing Time
| style="background-color:#9aff99;" | Processing Time
| style="background-color:#9aff99;" | FPS
| style="background-color:#9aff99;" | FPS
| style="background-color:#abe9ff;" | CPU%
| style="background-color:#96fffb;" | CPU%
| style="background-color:#abe9ff;" | Processing Time
| style="background-color:#96fffb;" | Processing Time
| style="background-color:#abe9ff;" | FPS
| style="background-color:#96fffb;" | FPS
| style="background-color:#abe9ff;" | CPU%
| style="background-color:#96fffb;" | CPU%
| style="background-color:#abe9ff;" | Processing time
| style="background-color:#96fffb;" | Processing time
| style="background-color:#abe9ff;" | FPS
| style="background-color:#96fffb;" | FPS
| style="background-color:#ffffc7;" | CPU%
| style="background-color:#fffc9e;" | CPU%
| style="background-color:#ffffc7;" | Processing Time
| style="background-color:#fffc9e;" | Processing Time
| style="background-color:#ffffc7;" | FPS
| style="background-color:#fffc9e;" | FPS
| style="background-color:#ffffc7;" | CPU%
| style="background-color:#fffc9e;" | CPU%
| style="background-color:#ffffc7;" | Processing time
| style="background-color:#fffc9e;" | Processing time
| style="background-color:#ffffc7;" | FPS
| style="background-color:#fffc9e;" | FPS
|-
|-
| Value
| Value
Line 373: Line 373:
| style="background-color:#9aff99;" | 203.679 ms
| style="background-color:#9aff99;" | 203.679 ms
| style="background-color:#9aff99;" | 4.9
| style="background-color:#9aff99;" | 4.9
| style="background-color:#abe9ff;" | All: 3.35%<br />Core: 26.77%
| style="background-color:#96fffb;" | All: 3.35%<br />Core: 26.77%
| style="background-color:#abe9ff;" | 11.672 ms
| style="background-color:#96fffb;" | 11.672 ms
| style="background-color:#abe9ff;" | 85.68
| style="background-color:#96fffb;" | 85.68
| style="background-color:#abe9ff;" | All: 3.26%<br />Core: 26.05%
| style="background-color:#96fffb;" | All: 3.26%<br />Core: 26.05%
| style="background-color:#abe9ff;" | 11.459 ms
| style="background-color:#96fffb;" | 11.459 ms
| style="background-color:#abe9ff;" | 87.27
| style="background-color:#96fffb;" | 87.27
| style="background-color:#ffffc7;" | All: 11.32%<br />Core: 90.55%
| style="background-color:#fffc9e;" | All: 11.32%<br />Core: 90.55%
| style="background-color:#ffffc7;" | 363.230 ms
| style="background-color:#fffc9e;" | 363.230 ms
| style="background-color:#ffffc7;" | 2.75
| style="background-color:#fffc9e;" | 2.75
| style="background-color:#ffffc7;" | All: 11.42%<br />Core: 91.39%
| style="background-color:#fffc9e;" | All: 11.42%<br />Core: 91.39%
| style="background-color:#ffffc7;" | 359.669 ms
| style="background-color:#fffc9e;" | 359.669 ms
| style="background-color:#ffffc7;" | 2.78
| style="background-color:#fffc9e;" | 2.78
|}
|}

Revision as of 20:49, 19 December 2023

The library has two major components: detection of the desired object done by ONNX model and the application of the redaction effect done either by the CPU or GPU.

For testing purposes, take into account the folowing points:

  • Jetpack 5.1.X
  • Tested with GPU+CPU and CPU only. For GPU, execution done with TensorRT and CUDA.
  • The library offers two approaches: each method used separately and all together using the apply method.
  • The metrics measured are: CPU and GPU consumption, processing time and FPS.
  • The tracker is implemented only with support for CPU.

Xavier NX

In this section you will find the performance for the methods separated and for all of them together regarding the Jetson Xavier NX.

GPU

  • TensorRT
Image 1080p- 7 faces
Platform Xavier NX(6-core 20WMode )
Method Convert Detect Redact
Stat CPU% GPU%/VRAM Processing time FPS CPU% GPU%/VRAM Processing time FPS CPU% GPU%/VRAM Processing time FPS
Value All: 1.04%
Core: 6.25%
1.08%
85 MiB
4.988 ms 200.5 All: 4.42%
Core: 20.53%
3.37%
85 MiB
23.628 ms 42 All: 0.68%
Core: 4.11%
4.38%
85 MiB
20.024 ms 49.9
Image 1080p- 7 faces
Platform Xavier NX(6-core 20WMode )
Method Apply
Stat CPU% GPU%/VRAM Processing time FPS
Value All: 3.58%
Core: 21.46%
4.08%
85 MiB
33.3 ms 30
  • CUDA
Image 1080p- 7 faces
Platform Xavier NX(6-core 20WMode )
Method Convert Detect Redact
Stat CPU% GPU%/VRAM Processing time FPS CPU% GPU%/VRAM Processing time FPS CPU% GPU%/VRAM Processing time FPS
Value All: 1.05%
Core: 6.3%
1.039%
78 MiB
5 ms 200 All: 5.03%
Core: 30.17%
4.46%
83 MiB
27.78 ms 36 All: 0.8%
Core: 4.83%
4.8%
83 MiB
20.307 ms 49.2
Image 1080p- 7 faces
Platform Xavier NX(6-core 20WMode )
Method Apply
Stat CPU% GPU%/VRAM Processing time FPS
Value All: 5.41%
Core: 32.43%
4.68%
83 MiB
35.851 ms 27.9

CPU

Image 1080p- 7 faces
Platform Xavier NX(Desktop Mode )
Method Convert Detect Redact
Num threads 1 4 1 4 1 4
Stat CPU% Processing time FPS CPU% Processing time FPS CPU% Processing time FPS CPU% Processing Time FPS CPU% Processing Time FPS CPU% Processing time FPS
Value All: 23.36%
Core: 140.17%
2.738 ms 365 All: 23.33%
Core: 139.97%
2.723 ms 367 All: 13.67%
Core: 82.04%
157.073 ms 6.36 All: 41.42%
Core: 248.51%
94.265 ms 10.61 All: 14.68%
Core: 88.07%
239.640 ms 4.17 All: 14.67%
Core: 88.04%
236.452 ms 4.23
Image 1080p- 7 faces
Platform Xavier NX(Desktop Mode )
Method Apply
Num threads 1 4
Stat CPU% Processing time FPS CPU% Processing time FPS
Value All: 17.13%
Core: 102.77%
395,855 ms 2,53 All: 27.16%
Core: 162.94%
331,721 ms 3

Xavier AGX

In this section you will find the performance for the methods separated and for all of them together regarding the Jetson Xavier NX.

CPU

Image 1080p- 7 faces
Platform Xavier AGX(Desktop Mode )
Method Convert Detect Track Redact
Num threads 1 4 1 4 1 4 1 4
Stat CPU% Processing time FPS CPU% Processing time FPS CPU% Processing time FPS CPU% Processing Time FPS CPU% Processing Time FPS CPU% Processing time FPS CPU% Processing Time FPS CPU% Processing time FPS
Value All: 9.12%
Core: 9.12%
3,069 ms 325.84 All: 9.05%
Core: 36.21%
2,968 ms 336.93 All: 18.51%
Core: 74.03%
99,577 ms 10,04 All: 18.72%
Core: 74.86%
103,367 ms 9.67 All: 6.71%
Core: 26.83%
11,889 ms 84,11 All: 6.83%
Core: 27.32%
12,093 ms 82,69 All: 21.35%
Core: 85.38%
195,167 ms 5,12 All: 21.31%
Core: 85.25%
194,899 ms 5,12
Image 1080p- 7 faces
Platform Xavier AGX(8-cores 30W Mode )
Method Convert Detect Track Redact
Num threads 1 4 1 4 1 4 1 4
Stat CPU% Processing time FPS CPU% Processing time FPS CPU% Processing time FPS CPU% Processing Time FPS CPU% Processing Time FPS CPU% Processing time FPS CPU% Processing Time FPS CPU% Processing time FPS
Value All: 5.95%
Core: 47.57%
2.075 ms 481.93 All: 5.88%
Core: 47.04%
2.189 ms 456.83 All: 10.52%
Core: 84.14%
204.358 ms 4.89 All: 10.51%
Core: 84.09%
203.679 ms 4.9 All: 3.35%
Core: 26.77%
11.672 ms 85.68 All: 3.26%
Core: 26.05%
11.459 ms 87.27 All: 11.32%
Core: 90.55%
363.230 ms 2.75 All: 11.42%
Core: 91.39%
359.669 ms 2.78