- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
hello
in another discussion ,i know that if i want to infer a image and get the num of item in this image,i need ssd or yolo model,so i go to TensorFlow ,and then i got a ssd_mobilenet_v1_coco_2018_01_28.tar.gz , i need to compile it to graph file for my movidius,I'm successful , i got graph and labels file, but when i use these files to my program , it always get error back : output ,userobj = fifoout.read_elem() ,output[0] is 0 . i think this graph only result one item like inception_v1
so i want to know , did i use right model? if not ,can you show me how to get it?
is there anyone have same problem? how you fix it?
this is where model from
this is how to compile
https://movidius.github.io/ncsdk/tf_modelzoo.html
thank you
- Tags:
- Caffe
- Tensorflow
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
i checked api again , and i found movidius can not use TensorFlow ssd_mobilenet , i'm going to try make TensorFlow model to caffe model…

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page