peteryuX / retinaface-tf2

RetinaFace (Single-stage Dense Face Localisation in the Wild, 2019) implemented (ResNet50, MobileNetV2 trained on single GPU) in Tensorflow 2.0+. This is an unofficial implementation. With Colab.

peteryuX/retinaface-tf2 Issues

Size mismatch error which running mobilenet model
Closed a year ago2
how convert to .tflite?
Updated 7 months ago5
Error Test.py on one image
Updated 8 months ago7
Error in converting retinaface mobilenetv2 backend to CoreML
Updated a year ago
ValueError: Received incompatible tensor with shape (1, 1, 32, 64) when attempting to restore variable with shape (1, 1, 192, 64) and name model/layer_with_weights-1/output1/conv/kernel/.ATTRIBUTES/VARIABLE_VALUE.
Closed 2 years ago5
How to train with custom dataset by using the pretrained model?
Updated a year ago
Cannot find ckpt from ./checkpoints/retinaface_mbv2.
Closed 2 years ago1
Error Testing on WIDER FACE Validation Set
Closed 2 years ago1
Access of widerface
Updated 2 years ago
Empty output on second inference call on exported saved model
Closed 2 years ago1
SSH context module
Updated 2 years ago1
Error while running on tf2.4
Updated 2 years ago3
frozen graph
Updated 3 years ago
I'm wondering how to fine-tune the model, such as freezing the model, which levels can have a good training effect. Is there any research?
Updated 3 years ago
PiecewiseConstantWarmUpDecay incorrect implementation causes skipping to the almost minimal LR
Closed 3 years ago
More than 5 landmarks
Closed 3 years ago1
test performance on big face
Updated 3 years ago1
train size
Updated 3 years ago
Very slow inference
Updated 3 years ago2
How can i saved pretrained model as "savedmodel" tensorflow format ?
Updated 3 years ago9
Inference with TF1.15
Closed 4 years ago2
OSError: [WinError 126] The specified module could not be found
Updated 3 years ago
keras .h5 model
Updated 3 years ago
inference time
Updated 3 years ago
the question about the inference time
Updated 3 years ago1
URL fetch failed
Closed 3 years ago1
tf.checkpoint how to convert to pb correct
Closed 4 years ago9
Can you add inference script?
Updated 4 years ago8
classification loss (crossentropy)
Updated 4 years ago
CPU RAM consumption until depletion during training if shuffle is set to True
Closed 4 years ago1
Could you provide a pre-training model suitable for tensorflow1.10?
Updated 4 years ago
Completely wrong prediction matrix!
Updated 4 years ago
image input shape does not have to be the same as Input?
Closed 4 years ago1
GPU utilisation is 0 while running test.py
Updated 4 years ago
support batching on test
Updated 4 years ago
about multi gpus
Closed 4 years ago2
Typo in the README
Closed 4 years ago1
hard negative mining
Closed 4 years ago2
can not find esrgan.yaml and --save_image in test.py
Closed 4 years ago2
Runtime performance
Closed 4 years ago1
Can this tf2 model be converted into pb model like in tensorflow 1.x and then be executed in tensorflow 1.x ?
Closed 4 years ago1
Pre model download
Closed 4 years ago4