about the config of the mobiledet ssdlite
Keval-WOT opened this issue · 6 comments
Hello There , I want to train SSDLite Mobiledet model for my custom dataset , Can you please help me where can i find its config and also if possible can you please share colab or something like that to train the model
My task is to train custom SSDLite mobiledet model and run it on my coral edge device
Hi @Keval-WOT ,
Here are the configuration for SSDLite Mobiledet config but i don't think models/research has any colab notebook tutorials. But official/models has ample of notebooks for object detection on how to do custom training and evaluation. If you want to have look at it , please check these Object detection tutorials.Thanks
@laxmareddyp thanks for a quick response , I would like to ask you one more question that basically i want to train a person detection model on cctv feed so just want to know that which model should be best for that particular task since i also need to handle multiple stream with a decent fps so which model should be best while balancing the accuracy and speed on the edge tpu M2 board with dual coral TPU
Hi @Keval-WOT,
For person detection on CCTV feeds with multiple streams and a decent frame rate, while balancing accuracy and speed on the Edge TPU M2 board with dual Coral TPUs, the YOLO series models would be a good choice at first place.While other models like RetinaNet ,EfficientDet-Lite,or Faster R-CNN could also be considered. However, it's always recommended to evaluate multiple models and configurations on your specific hardware and dataset to determine the best fit for your requirements.
Thanks.
This issue has been marked stale because it has no recent activity since 7 days. It will be closed if no further activity occurs. Thank you.
This issue was closed due to lack of activity after being marked stale for past 7 days.