-
Do you have plans to add MobilenetV2.cfg to yolo?
Link to article:
https://arxiv.org/pdf/1801.04381.pdf
Quote from the article:
> MobileNetV2 SSDLite is 20× more efficient and
> 10× smaller w…
-
Hello, I am trying to prune the backbone of an object detection model, specifically:
mmdet::configs/ssd/ssdlite_mobilenetv2-scratch_8xb24-600e_coco.py
General information on pruning image classifi…
-
https://github.com/open-mmlab/mmdetection/blob/main/configs/ssd/ssdlite_mobilenetv2-scratch_8xb24-600e_coco.py
I want the in_channel of mobilenet for that model to be 4 because the data I will be i…
-
Hi guys.
I did every steps in https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/running_on_mobile_tensorflowlite.md
but I can't n my FLOAT mobilenetv2 ssdlite on andr…
-
Hello,
Are you using pre-trained weights to initialize your training? I am referring in particular to Mobilenetv2-ssdlite. If yes, where can I find these pre-trained weights?
Thank you,
-
Thank's for a great tutorial!
I have tried to do a similar thing (with fruits dataset) but with the following models (TF1 OD model zoo) ssdlite_mobilenet_v2_coco_2018_05_09 and ssd_inception_v2_coco_…
-
Thanks for this Great work. I notice that the input size of mobilenet-SSDlite is 320x320 in the original paper, wheras it is 300x300 in the deploy.prototxt. Why are they different?
-
Awesome idea of execution.
After following all the steps, when i execute the demo_caffe.py project, facing below issue
I0530 12:31:54.303753 14700 layer_factory.hpp:77] Creating layer Conv/relu
F05…
-
Hi, when I run:
python3 model_main.py --logtostderr --train_dir=training/ --pipeline_config_path=training/ssdlite_mobilenet_v2_coco.config
I get this issue:
self._session = tf_session.TF_NewSessi…
-
I used tensorflow object detection api to train MobileNetV3,https://github.com/tensorflow/models/blob/master/research/object_detection/samples/configs/ssdlite_mobilenet_v3_large_320x320_coco.config, A…