AliButtar commented 4 years ago

Prerequisites

Please answer the following questions for yourself before submitting an issue.

[X] I am using the latest TensorFlow Model Garden release and TensorFlow 2.
[X] I am reporting the issue to the correct repository. (Model Garden official or research directory)
[X] I checked to make sure that this issue has not already been filed.

1. The entire URL of the file you are using

The notebook: Object Detection Tutorial Notebook
The model used: http://download.tensorflow.org/models/object_detection/tf2/20200711/mask_rcnn_inception_resnet_v2_1024x1024_coco17_gpu-8.tar.gz
The colab notebook that uses the same notebook and the model: Colab Object Detection Tutorial Copy

2. Describe the bug

Error Name : InvalidArgumentError

Error Description : Index out of range using input dim 1; input has only 1 dims [Op:StridedSlice] name: strided_slice/

The code in the Object Detection Tutorial Notebook successfully infers using object detection models but fails to infer using using Mask RCNN Inception Model and causes the above error.

3. Steps to reproduce

Run the following Colab Notebook. It is same as the Object Detection Tutorial but is instead using Mask RCNN Inception Resnet v2 1024x1204 model.

Colab Object Detection Tutorial Copy

4. Expected behavior

Segmentation on the test images same the as the object detection on the test images in the notebook

5. Additional context

The notebook crashes on the default segmentation model used in the Object Detection Tutorial Notebook as well also models that have been further trained to segment or detect other objects

6. System information

All the work done was on Colab with GPU enabled

ravikyram commented 4 years ago

I have tried in colab with Tf version 2.3 and was able to reproduce the issue.Please, find the gist here. Thanks!

jersonal commented 4 years ago

i get the same Error,did you solve this Error？

jersonal commented 4 years ago

i get the same Error,did you solve this Error？

i solved it. it because 'image_shape' and 'num_proposals' node are the parameters that needs to be feed.so dont use them,and convert to tensor

AliButtar commented 4 years ago

i get the same Error,did you solve this Error？

i solved it. it because 'image_shape' and 'num_proposals' node are the parameters that needs to be feed.so dont use them,and convert to tensor

Can you go into a little more detail, and share the solution perhaps. It will help a lot. Thanks

jersonal commented 4 years ago

i get the same Error,did you solve this Error？

i solved it. it because 'image_shape' and 'num_proposals' node are the parameters that needs to be feed.so dont use them,and convert to tensor

Can you go into a little more detail, and share the solution perhaps. It will help a lot. Thanks

Change "run_inference_for_single_image" function `def run_inference_for_single_image(model, image):

image = np.asarray(image)
# The input needs to be a tensor, convert it using `tf.convert_to_tensor`.
input_tensor = tf.convert_to_tensor(image)
# The model expects a batch of images, so add an axis with `tf.newaxis`.
input_tensor = input_tensor[tf.newaxis, ...]
# Run inference
model_fn = model.signatures['serving_default']
output_dict = model_fn(input_tensor)
# All outputs are batches tensors.
# Convert to numpy arrays, and take index [0] to remove the batch dimension.
# We're only interested in the first num_detections.
num_detections = int(output_dict.pop('num_detections'))
need_detection_key = ['detection_classes','detection_boxes','detection_masks','detection_scores']
output_dict = {key: output_dict[key][0, :num_detections].numpy()
               for key in need_detection_key}
output_dict['num_detections'] = num_detections
# detection_classes should be ints.
output_dict['detection_classes'] = output_dict['detection_classes'].astype(np.int64)
# Handle models with masks:
if 'detection_masks' in output_dict:
    # Reframe the the bbox mask to the image size.
    detection_masks_reframed = utils_ops.reframe_box_masks_to_image_masks(
        tf.convert_to_tensor(output_dict['detection_masks']), output_dict['detection_boxes'],
        image.shape[0], image.shape[1])
    detection_masks_reframed = tf.cast(detection_masks_reframed > 0.5,
                                       tf.uint8)
    output_dict['detection_masks_reframed'] = detection_masks_reframed.numpy()

return output_dict

`

AliButtar commented 4 years ago

i get the same Error,did you solve this Error？

i solved it. it because 'image_shape' and 'num_proposals' node are the parameters that needs to be feed.so dont use them,and convert to tensor

Can you go into a little more detail, and share the solution perhaps. It will help a lot. Thanks

Change "run_inference_for_single_image" function `def run_inference_for_single_image(model, image):

image = np.asarray(image)
# The input needs to be a tensor, convert it using `tf.convert_to_tensor`.
input_tensor = tf.convert_to_tensor(image)
# The model expects a batch of images, so add an axis with `tf.newaxis`.
input_tensor = input_tensor[tf.newaxis, ...]
# Run inference
model_fn = model.signatures['serving_default']
output_dict = model_fn(input_tensor)
# All outputs are batches tensors.
# Convert to numpy arrays, and take index [0] to remove the batch dimension.
# We're only interested in the first num_detections.
num_detections = int(output_dict.pop('num_detections'))
need_detection_key = ['detection_classes','detection_boxes','detection_masks','detection_scores']
output_dict = {key: output_dict[key][0, :num_detections].numpy()
               for key in need_detection_key}
output_dict['num_detections'] = num_detections
# detection_classes should be ints.
output_dict['detection_classes'] = output_dict['detection_classes'].astype(np.int64)
# Handle models with masks:
if 'detection_masks' in output_dict:
    # Reframe the the bbox mask to the image size.
    detection_masks_reframed = utils_ops.reframe_box_masks_to_image_masks(
        tf.convert_to_tensor(output_dict['detection_masks']), output_dict['detection_boxes'],
        image.shape[0], image.shape[1])
    detection_masks_reframed = tf.cast(detection_masks_reframed > 0.5,
                                       tf.uint8)
    output_dict['detection_masks_reframed'] = detection_masks_reframed.numpy()

return output_dict

`

Thank You so much, it worked

ravikyram commented 4 years ago

@AliButtar

Please, close this thread if your issue was resolved. Thanks!

google-ml-butler[bot] commented 4 years ago

Are you satisfied with the resolution of your issue? Yes No

Petros626 commented 3 years ago

i get the same Error,did you solve this Error？

i solved it. it because 'image_shape' and 'num_proposals' node are the parameters that needs to be feed.so dont use them,and convert to tensor

Can you go into a little more detail, and share the solution perhaps. It will help a lot. Thanks

Change "run_inference_for_single_image" function `def run_inference_for_single_image(model, image):

image = np.asarray(image)
# The input needs to be a tensor, convert it using `tf.convert_to_tensor`.
input_tensor = tf.convert_to_tensor(image)
# The model expects a batch of images, so add an axis with `tf.newaxis`.
input_tensor = input_tensor[tf.newaxis, ...]
# Run inference
model_fn = model.signatures['serving_default']
output_dict = model_fn(input_tensor)
# All outputs are batches tensors.
# Convert to numpy arrays, and take index [0] to remove the batch dimension.
# We're only interested in the first num_detections.
num_detections = int(output_dict.pop('num_detections'))
need_detection_key = ['detection_classes','detection_boxes','detection_masks','detection_scores']
output_dict = {key: output_dict[key][0, :num_detections].numpy()
               for key in need_detection_key}
output_dict['num_detections'] = num_detections
# detection_classes should be ints.
output_dict['detection_classes'] = output_dict['detection_classes'].astype(np.int64)
# Handle models with masks:
if 'detection_masks' in output_dict:
    # Reframe the the bbox mask to the image size.
    detection_masks_reframed = utils_ops.reframe_box_masks_to_image_masks(
        tf.convert_to_tensor(output_dict['detection_masks']), output_dict['detection_boxes'],
        image.shape[0], image.shape[1])
    detection_masks_reframed = tf.cast(detection_masks_reframed > 0.5,
                                       tf.uint8)
    output_dict['detection_masks_reframed'] = detection_masks_reframed.numpy()

return output_dict

`

hey i use the code you shared with us, but in run in new errors:

github1 github2

518789Adf commented 3 months ago

input has only 1 dims for '{{node strided_slice_8}}

Petros626 commented 3 months ago

input has only 1 dims for '{{node strided_slice_8}}

Look at my mentioned issue

tensorflow / models

InvalidArgumentError: Index out of range using input dim 1; input has only 1 dims [Op:StridedSlice] name: strided_slice/ #9255

Prerequisites

1. The entire URL of the file you are using

2. Describe the bug

3. Steps to reproduce

4. Expected behavior

5. Additional context

6. System information