Update icon, colors object detection

[st-tuanmai]

No description provided.

[PaulTR]

Can you verify the run script for downloading the models? Getting

ptruiz-macbookpro:ios ptruiz$ sh RunScripts/download_models.sh
+++ dirname RunScripts/download_models.sh
++ cd RunScripts
++ pwd

• SCRIPT_DIR=/Users/ptruiz/Documents/code/ai-edge/interpreter-samples/examples/object_detection/ios/RunScripts
• MOBILENETV1_SSD_URL=https://storage.googleapis.com/download.tensorflow.org/models/tflite/task_library/object_detection/ios/lite-model_ssd_mobilenet_v1_1_metadata_2.tflite
• EFFICIENTDET_LITE0_URL=https://storage.googleapis.com/download.tensorflow.org/models/tflite/task_library/object_detection/ios/lite-model_efficientdet_lite0_detection_metadata_1.tflite
• EFFICIENTDET_LITE1_URL=https://storage.googleapis.com/download.tensorflow.org/models/tflite/task_library/object_detection/ios/lite-model_efficientdet_lite1_detection_metadata_1.tflite
• EFFICIENTDET_LITE2_URL=https://storage.googleapis.com/download.tensorflow.org/models/tflite/task_library/object_detection/ios/lite-model_efficientdet_lite2_detection_metadata_1.tflite
• MOBILENETV1_SSD_NAME=ssd_mobilenet_v1.tflite
• EFFICIENTDET_LITE0_NAME=efficientdet_lite0.tflite
• EFFICIENTDET_LITE1_NAME=efficientdet_lite1.tflite
• EFFICIENTDET_LITE2_NAME=efficientdet_lite2.tflite
  ++ mktemp -d
• DOWNLOADS_DIR=/var/folders/m3/71gcw4c545g0cpkjhx954kqm00nvwc/T/tmp.oK9kLDxWGe
• cd /Users/ptruiz/Documents/code/ai-edge/interpreter-samples/examples/object_detection/ios/RunScripts
• has_download=false
• '[' -f ../ObjectDetection/TFLite/ssd_mobilenet_v1.tflite ']'
• echo 'File ssd_mobilenet_v1.tflite exists.'
  File ssd_mobilenet_v1.tflite exists.
• '[' -f ../ObjectDetection/TFLite/efficientdet_lite0.tflite ']'
• echo 'File efficientdet_lite0.tflite exists.'
  File efficientdet_lite0.tflite exists.
• '[' -f ../ObjectDetection/TFLite/efficientdet_lite1.tflite ']'
• echo 'File efficientdet_lite1.tflite exists.'
  File efficientdet_lite1.tflite exists.
• '[' -f ../ObjectDetection/TFLite/efficientdet_lite2.tflite ']'
• echo 'File efficientdet_lite2.tflite exists.'
  File efficientdet_lite2.tflite exists.
• false

[st-tuanmai]

Hi Paul, I check on my Mac, scripts work correctly. You don't need to run scripts manually, scripts will run automatically when building the app.

[PaulTR]

It also does not work from the IDE. Looking at it, the paths are likely wrong ("TFLite"), and the storage link for the tflite files are incorrect. There should only be efficientdet 0 and 2 and they should come from the ai-edge bucket: https://storage.googleapis.com/ai-edge/interpreter-samples/object_detection/ios/efficientdet_lite0.tflite and https://storage.googleapis.com/ai-edge/interpreter-samples/object_detection/ios/efficientdet_lite2.tflite

Please use image classification download_models.sh as your template.

[st-tuanmai]

I tried deleting the model and running it again, it still works fine. I will check again.

[st-tuanmai]

The script works fine on my mac.

[PaulTR]

Again, it's still using the wrong model paths, bringing in more models than needed, and I'd like it to follow the same format as the other examples.

[st-tuanmai]

Sorry @PaulTR
I missed some messages

[st-tuanmai]

The output of the new model is not the same structure as the old model so it cannot be fixed quickly, I will try to fix it tomorrow.

[st-tuanmai]

@PaulTR
I don't have a way to map the boxes of this new model to the original image, can you tell me how this model is trained, or is there a python sample to use, or a dataset to refer to?
Thank you.

[PaulTR]

See if this tool helps, otherwise I'll check it over on Tuesday https://github.com/google-ai-edge/model-explorer

…

On Sat, Nov 2, 2024, 9:48 PM Tuan Mai A. ***@***.***> wrote: @PaulTR <https://github.com/PaulTR> I don't have a way to map the boxes of this new model to the original image, can you tell me how this model is trained, or is there a python sample to use, or a dataset to refer to? Thank you. — Reply to this email directly, view it on GitHub <#21 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAR2C345W47QJVM5QEJEUSTZ6WMKNAVCNFSM6AAAAABNV5E3KWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINJTGI4DENJYHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[PaulTR]

Did you have any luck with this using the model links I sent?