AxisCommunications / acap-native-sdk-examples

Example code for APIs and features in AXIS Camera Application Platform (ACAP) Native SDK
Apache License 2.0
41 stars 23 forks source link

Inference is performed, but no object is detected.(object_detection-cv25) #141

Closed tosamu51 closed 1 year ago

tosamu51 commented 1 year ago

Describe the bug

Inference is performed, but no object is detected.(object_detection-cv25)

To reproduce

Run object-detection-cv25 in M2035-LE and M3085-V ,Inference is performed, but no object is detected.

object_detection :detect object Person,Chair... (M2035_ObjectDetect.txt) https://github.com/AxisCommunications/acap-native-sdk-examples/tree/main/object-detection

object_detection-cv25 :Not detect object (M2035_ObjectDetectCV25.txt) https://github.com/AxisCommunications/acap-native-sdk-examples/tree/main/object-detection-cv25

Camera1:AXIS M2035-LE(11.5.64) Camera2:AXIS M3085-V(11.5.64) AXIS M2035-LE and AXIS M3085-V(11.5.64) is same phenomenon

Log

2023-09-04T04:38:54.927+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Starting /usr/local/packages/object_detection/object_detection 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'buffer.strategy': <uint32 3> 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'channel'--------: <uint32 1> 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'format'---------: <uint32 3> 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'height'---------: <uint32 720> 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'width'----------: <uint32 1280> 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Creating VDO image provider and creating stream 1280 x 720 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Dump of vdo stream settings map ===== 2023-09-04T04:38:55.258+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: chooseStreamResolution: We select stream w/h=1280 x 720 based on VDO channel info. 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'buffer.strategy': <uint32 3> 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'channel'--------: <uint32 1> 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'format'---------: <uint32 3> 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'height'---------: <uint32 360> 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: 'width'----------: <uint32 640> 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Creating VDO raw image provider and stream 640 x 360 2023-09-04T04:38:55.644+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Dump of vdo stream settings map ===== 2023-09-04T04:38:55.820+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: raw width=640, raw height=360 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Calculate crop image 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Create larod models 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Create preprocessing maps 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Creating VDO crop image provider and creating stream 1280 x 720 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Crop VDO image X=280 Y=0 (720 x 720) 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Setting up larod connection with chip ambarella-cvflow, model /usr/local/packages/object_detection/model/converted_model.bin and label file /usr/local/packages/object_detection/label/labels.txt 2023-09-04T04:38:55.879+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: chooseStreamResolution: We select stream w/h=1280 x 720 based on VDO channel info. 2023-09-04T04:38:55.911+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Available chip IDs: 2023-09-04T04:38:55.911+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Chip: ambarella-cvflow 2023-09-04T04:38:55.911+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Chip: ambarella-cvflow-proc 2023-09-04T04:38:55.911+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Chip: cpu-proc 2023-09-04T04:38:55.911+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Chip: cpu-tflite 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Allocate memory for input/output buffers 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Create input/output tensors 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Creating temporary files and memmaps for inference input and output tensors 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Determine tensor buffer sizes 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Loading preprocessing model with chip cpu-proc 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: createAndMapTmpFile: Setting up a temp fd with pattern /tmp/crop.test-XXXXXX and size 691200 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: createAndMapTmpFile: Setting up a temp fd with pattern /tmp/larod.in.test-XXXXXX and size 288000 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: createAndMapTmpFile: Setting up a temp fd with pattern /tmp/larod.out1.test-XXXXXX and size 30672 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: createAndMapTmpFile: Setting up a temp fd with pattern /tmp/larod.out2.test-XXXXXX and size 697788 2023-09-04T04:38:56.110+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: createAndMapTmpFile: Setting up a temp fd with pattern /tmp/larod.pp.test-XXXXXX and size 1382400 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Connect tensors to file descriptors 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Create job requests 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Found 1 input tensors and 2 output tensors 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Set input tensors 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Set output tensors 2023-09-04T04:38:56.114+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Start fetching video frames from VDO 2023-09-04T04:38:56.128+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 22 ms 2023-09-04T04:38:56.157+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 27 ms 2023-09-04T04:38:56.257+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 101 ms 2023-09-04T04:38:56.271+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 13 ms 2023-09-04T04:38:56.298+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 26 ms 2023-09-04T04:38:56.399+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 100 ms 2023-09-04T04:38:56.414+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 15 ms 2023-09-04T04:38:56.441+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 26 ms 2023-09-04T04:38:56.544+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 102 ms 2023-09-04T04:38:56.559+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 15 ms 2023-09-04T04:38:56.586+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 26 ms 2023-09-04T04:38:56.686+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 99 ms 2023-09-04T04:38:56.699+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 13 ms 2023-09-04T04:38:56.726+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 26 ms 2023-09-04T04:38:56.828+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 101 ms 2023-09-04T04:38:56.842+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 14 ms 2023-09-04T04:38:56.870+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 27 ms 2023-09-04T04:38:56.969+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 99 ms 2023-09-04T04:38:56.986+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 16 ms 2023-09-04T04:38:57.014+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 28 ms 2023-09-04T04:38:57.115+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Postprocesing in 99 ms 2023-09-04T04:38:57.127+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Converted image in 13 ms 2023-09-04T04:38:57.154+00:00 axis-b8a44f45d2c1 [ INFO ] object_detection[1292144]: Ran inference for 26 ms

Environment

M2035_ObjectDetectCV25.txt M3085_ObjectDtectCV25.txt M2035_ObjectDetect.txt

Corallo commented 1 year ago

Hello @tosamu51

We are aware of this issue, we are currently investigating the problem. We'll get back to you when we have an update.

pataxis commented 1 year ago

Hi @tosamu51 , this issue should have been fixed and has been synced to main branch. Could you retest?

tosamu51 commented 1 year ago

Thanks for the fix. We have confirmed that it works with AXIS M2035-LE (11.5.64).

object_detectionLOG.txt