Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Employee
161 Views

Videos as input to OpenVINO benchmark python tool

Jump to solution

Is there a way we can give video as input file (-i) to python benchmark tool instead of images.  https://docs.openvinotoolkit.org/latest/openvino_inference_engine_tools_benchmark_tool_README.html 

0 Kudos

Accepted Solutions
Highlighted
Moderator
121 Views

Hi Swanand,


Video input is not supported for Benchmark Python Tool. For a particular model, use Benchmark Python Tool, using different devices, e.g. CPU, GPU and VPU, different batch sizes, sync or async execution, different number of infer requests or threads, and so on, to find the best configuration that optimizes your inference.


For video input, I would suggest you try using the relevant demos or samples that support video inputs. For example, Interactive Face Detection Demo for object detection, Smart Classroom Demo for action recognition, and Pedestrian Tracker Demo for person detection and recognition, etc.


Regards,

Munesh


View solution in original post

0 Kudos
2 Replies
Highlighted
Moderator
122 Views

Hi Swanand,


Video input is not supported for Benchmark Python Tool. For a particular model, use Benchmark Python Tool, using different devices, e.g. CPU, GPU and VPU, different batch sizes, sync or async execution, different number of infer requests or threads, and so on, to find the best configuration that optimizes your inference.


For video input, I would suggest you try using the relevant demos or samples that support video inputs. For example, Interactive Face Detection Demo for object detection, Smart Classroom Demo for action recognition, and Pedestrian Tracker Demo for person detection and recognition, etc.


Regards,

Munesh


View solution in original post

0 Kudos
Highlighted
Moderator
102 Views

Hi Swanand,


This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.


Regards,

Munesh


0 Kudos