I want to implement YOLO-tiny in Keras with Tensorflow 2.0 backend. I want to make a new custom YoloLayer that performs non-maximum suppression to outputs of the previous layer and makes tensor with shape (batch_size, num, 6), where num is a number of found predictions, and each prediction presented as [x, y, w, h, prob, class]. I also set self.trainable = False in __init__() method. Here is my call method:
def call(self, inputs, **kwargs):
predictions = inputs[...,:5]
x = tf.math.add(self.cols, tf.nn.sigmoid(predictions[...,0])) / self.grid_size # x
y = tf.math.add(self.rows, tf.nn.sigmoid(predictions[...,1])) / self.grid_size # y
w = tf.multiply(self.anchors_w, tf.math.exp(predictions[...,2])) / self.grid_size # w
h = tf.multiply(self.anchors_h, tf.math.exp(predictions[...,3])) / self.grid_size # h
c = tf.nn.sigmoid(predictions[...,4]) # confidence
bounds = tf.stack([x, y, w, h], -1)
classes = inputs[...,5:]
probs = tf.multiply(tf.nn.softmax(classes), tf.expand_dims(c, axis=-1))
prob_mask = tf.greater(probs, self.threshold)
suppressed_indices = tf.where(prob_mask)
suppressed_probs = tf.gather_nd(probs, suppressed_indices[...,:3])
suppressed_boxes = tf.gather_nd(bounds, suppressed_indices[...,:3])
box_coords = tf.stack([
suppressed_boxes[...,1] - suppressed_boxes[...,3] / 2., #y1
suppressed_boxes[...,0] - suppressed_boxes[...,2] / 2., #x1
suppressed_boxes[...,1] + suppressed_boxes[...,3] / 2., #y2
suppressed_boxes[...,0] + suppressed_boxes[...,2] / 2., #x2
], axis=-1)
out = tf.TensorArray(tf.float32, size=0, dynamic_size=True)
for i in range(tf.shape(inputs)[0]):
image_out = tf.TensorArray(tf.float32, size=self.classes)
for c in range(self.classes):
class_probs = suppressed_probs[i,:,c]
indices = tf.image.non_max_suppression(box_coords[i], class_probs, 10,
iou_threshold=self.nms_threshold,
score_threshold=self.threshold)
if tf.size(indices) > 0:
final_probs = tf.expand_dims(tf.gather(class_probs, indices), axis=-1)
final_boxes = tf.gather(suppressed_boxes[i], indices)
class_vec = tf.ones((tf.shape(final_probs)[0], 1)) * c
image_out.write(c, tf.concat([final_boxes, final_probs, class_vec], axis=1))
image_out = image_out.concat()
out.write(i, image_out)
out = out.stack()
return out
Then, model.summary() returns:
Model: "sequential_1"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
...
_________________________________________________________________
yolo_layer (YoloLayer) (None, None, 6) 0
=================================================================
...
I loaded pre-trained weights for this model and ran model.predict, but the output gave me an error:
InvalidArgumentError: Tried to stack elements of an empty list with non-fully-defined element_shape: [?,6]
[[node sequential_1/yolo_layer/TensorArrayV2Stack/TensorListStack (defined at <ipython-input-2-fbae137dd1a2>:96) ]] [Op:__inference_predict_function_4604]
I also ran this model without YoloLayer and modified its outputs with the same function but separate, and it works correct, but it isn't with placeholders. What should I do to acheive this?
Ok, I found it out by myself. All had to be done is:
outputs = outputs.write(out_idx, image_out)