Stack Overflow
Stack Overflow

Questions tagged [mobilenet]

205 questions
0
votes
1 answer

Mobilenet SSD input shape

I have trained a mobile SSD V2 model on a custom dataset for a single object detection task, and I have converted it to tflite. When I load the .tflite model using the interpreter for testing, and geting the input details using : input_details =…
Marwen Bhj
  • 67
  • 5
0
votes
0 answers

How to set up pipeline.config file with one class to train on mobilenet v3 large? and also how to set fine tune checkpoint?

I have to set class of one in pipeline.config,i changed the number of classes to 1 and i started training its showing an error InvalidArgumentError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a mismatch…
Naveen Kumar
  • 1
0
votes
1 answer

Calculation operations with the parameters of a TFLite quantized model

I am trying to implement image classification in hardware using the quantized Mobilenetv2 model taken from here. To do that, I first need to reproduce the inference process from the beginning to the end to make sure I understand the…
Nazar
  • 820
  • 5
  • 13
  • 36
0
votes
1 answer

Problem with MobilenetV2 implementation on mobile

MobilenetV2 tflite model trained on cifar10 dataset is expecting input image of shape 1x32x32x3(Because during training their was batch flag used) but in order to dump model into Mobile I get 32x32x3 shape images from camera so now not able to…
Akash Bhogar
  • 1
  • 1
  • 1
0
votes
1 answer

ssd_mobilenet_v1_coco vs ssd_mobilenet_v1_quantized_coco

I know one is probably trained with quantization aware trained and is quantize while the other is not. Is there any difference in both of their checkpoints? because both have checkpoints of same size. I wanted to train ssd_mobilenet_v1 for my own…
MrKhan
  • 154
  • 12
0
votes
1 answer

I tried to train an image classifier based on MobilenetV2, but loss has not been able to converge, I am not sure if I use tensorflow correctly

I tried to use the MobilenetV2 model as an image classifier. There are 10 categories, which are randomly selected from ImageNet's classification dataset. But loss does not converge during training, it always tends to fluctuate back and forth with a…
chenbin tang
  • 61
  • 2
0
votes
1 answer

Unable to infer results using tflite object detection model

I successfully converted a ssd_mobilenet_v3 model into a .tflite by retraining the model. (previously trained on coco dataset) But while using this model for object detection to run inference on a single image using the following code: interpreter =…
Sushanth
  • 2,224
  • 13
  • 29
0
votes
1 answer

Training Keras MobileNetV2 on CIFAR-100 (from scratch)

I want to train MobileNetV2 from scratch on CIFAR-100 and I get the following results where it just stops learning after some while. Here is my code. I would like to see at least 60-70% validation accuracy and I wonder whether I have to pre-train…
Overholt
  • 867
  • 1
  • 10
  • 23
0
votes
2 answers

Not getting Proper Accurcacy for cifar10 dadatset with mobilenetv2 on CPU

I have tried to train mobilenetV2 with cifar10 dataset as mentioned in https://github.com/tensorflow/models/tree/master/research/slim only change was I did it on CPU but i ended getting only 63% accuracy but the stated accuracy is…
Akash Bhogar
  • 1
  • 1
  • 1
0
votes
1 answer

How to do incremental learning on MobileNet-SSD caffe

I'm training my classifier on 20k images but every week I get more new pictures so I want to incrementally train my previous model(last stopped iteration) instead of retraining 20k+new_images on all the images again which is a waste of time and…
abhimanyuaryan
  • 3,882
  • 5
  • 41
  • 83
0
votes
1 answer

How to add my own dataset to MobileNet model?

I am trying to do prediction by loading my own dataset to the MobileNet model.According to this tutorial dataset(cats_vs_dogs) is loaded from the TensorFlow dataset as shown bellow splits = tfds.Split.ALL.subsplit(weighted=(80, 20)) splits, info =…
rasindu alwis
  • 43
  • 1
  • 3
0
votes
0 answers

CNN classification with images with watermarks in dataset

I was wondering if having images with watermarks in the training set when training a CNN (MobileNetv2) will affect its accuracy. Would simply removing them from the set be a better thing to do? Thank you!
Russell C.
  • 588
  • 6
  • 25
0
votes
1 answer

try to use "load_model" but he creates new one with 0 layers

I want to load new fine tuned model with load_model. But Id dont load but instad creating a empty new one. The Model was build in the same environment with the same keras and Python versions. Sample codes do the same. model = load_weights is also…
Marv
  • 11
  • 3
0
votes
1 answer

Detecting small objects with MobileNet and Faster RCNN

I'm working on object detection of various sorts of animals using the Tensorflow Object Detection API. In the past I successfully applied MobileNet v1 to various settings and I used to be happy with the results. Now, I encountered a problem with a…
zimmermc
  • 623
  • 2
  • 6
  • 13
0
votes
1 answer

image size issue to train cifar dateset by standard mobileNet,VGG-16 or AlexNet from scratch in tensorflow or keras

Suppose,I want to train standard AlexNet, VGG-16 or MobileNet from scratch by CIFAR-10 or CIFAR-100 dataset in Tensorflow or Keras.Now the problem is that,the architecture of standard AlexNet,VGG-16 or MobileNet is built for ImageNet dataset where…
user2955839
  • 57
  • 1
  • 6
1 2 3
…
13
14