How can I use pretrained weights to predict

[engrsharoonhoney]

Code:
import cv2
import numpy as np
import LipNet
from tensorflow.keras.models import load_model

lipnet = LipNet()

videos_path = './data/s1/'
lipnet.model.load_weights("unseen-weights178.h5", by_name=True, skip_mismatch=True)

def process_video(video_path):
cap = cv2.VideoCapture(video_path)
frames = []

while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break
   
    gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
   
    resized_frame = cv2.resize(gray_frame, (100, 50))
   
    normalized_frame = resized_frame.astype(np.float32) / 255.0
   
    final_frame = np.expand_dims(normalized_frame, axis=-1)
   
    frames.append(final_frame)

cap.release()

frames_array = np.array(frames)

target_frames = 75
if frames_array.shape[0] < target_frames:
    padding = np.zeros((target_frames - frames_array.shape[0], 50, 100, 1), dtype=np.float32)
    frames_array = np.concatenate((frames_array, padding), axis=0)
elif frames_array.shape[0] > target_frames:
    frames_array = frames_array[:target_frames]

return np.expand_dims(frames_array, axis=0)

video_path = videos_path+"srbo2n.mpg"
video_frames = process_video(video_path)

print("Video Frames: ",video_frames.shape)

sentence_length = np.array([[video_frames.shape[1]]])
output_labels = np.zeros((1, 32)) # Placeholder label sequence
label_length = np.array([[32]]) # Placeholder label length

inputs = [video_frames, sentence_length, output_labels, label_length]

predicted_text = lipnet.model.predict_on_batch(inputs)

print("Predicted Speech:", predicted_text)

#------------------- Error ---------------------------------

Error:

Video Frames: (1, 75, 50, 100, 1)

ValueError Traceback (most recent call last)
Cell In[179], line 72
69 inputs = [video_frames, sentence_length, output_labels, label_length]
71 # Predict text from lip movements
---> 72 predicted_text = lipnet.model.predict_on_batch(inputs)
76 # Check the correct shape before passing to the model
77 # print("Final Input Shape:", video_frames.shape) # Should print (1, 75, 50, 100, 1)
78
79 # Predict text from lip movements
80 # predicted_text = lipnet.model.predict(video_frames)
81 print("Predicted Speech:", predicted_text)

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/backend/tensorflow/trainer.py:631, in TensorFlowTrainer.predict_on_batch(self, x)
629 def predict_on_batch(self, x):
630 self.make_predict_function()
--> 631 batch_outputs = self.predict_function([(x,)])
632 batch_outputs = tree.map_structure(
633 convert_to_np_if_not_ragged, batch_outputs
634 )
635 return batch_outputs

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/tensorflow/python/util/traceback_utils.py:153, in filter_traceback..error_handler(*args, **kwargs)
151 except Exception as e:
152 filtered_tb = _process_traceback_frames(e.traceback)
--> 153 raise e.with_traceback(filtered_tb) from None
154 finally:
155 del filtered_tb

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/backend/tensorflow/trainer.py:259, in TensorFlowTrainer.make_predict_function..one_step_on_data_distributed(data)
256 @tf.autograph.experimental.do_not_convert
257 def one_step_on_data_distributed(data):
258 data = data[0]
--> 259 outputs = self.distribute_strategy.run(
260 one_step_on_data, args=(data,)
261 )
262 outputs = reduce_per_replica(
263 outputs,
264 self.distribute_strategy,
265 reduction="concat",
266 )
267 return outputs

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/backend/tensorflow/trainer.py:249, in TensorFlowTrainer.make_predict_function..one_step_on_data(data)
246 @tf.autograph.experimental.do_not_convert
247 def one_step_on_data(data):
248 """Runs a predict test step on a batch of data."""
--> 249 return self.predict_step(data)

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/backend/tensorflow/trainer.py:104, in TensorFlowTrainer.predict_step(self, data)
102 x, _, _ = data_adapter_utils.unpack_x_y_sample_weight(data)
103 if self._call_has_training_arg:
--> 104 y_pred = self(x, training=False)
105 else:
106 y_pred = self(x)

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/utils/traceback_utils.py:122, in filter_traceback..error_handler(*args, **kwargs)
119 filtered_tb = _process_traceback_frames(e.traceback)
120 # To get the full stack trace, call:
121 # keras.config.disable_traceback_filtering()
--> 122 raise e.with_traceback(filtered_tb) from None
123 finally:
124 del filtered_tb

File ~/.config/jupyterlab-desktop/jlab_server/lib/python3.12/site-packages/keras/src/layers/input_spec.py:245, in assert_input_compatibility(input_spec, inputs, layer_name)
243 if spec_dim is not None and dim is not None:
244 if spec_dim != dim:
--> 245 raise ValueError(
246 f'Input {input_index} of layer "{layer_name}" is '
247 "incompatible with the layer: "
248 f"expected shape={spec.shape}, "
249 f"found shape={shape}"
250 )

ValueError: Input 0 of layer "functional_63" is incompatible with the layer: expected shape=(None, 75, 100, 50, 3), found shape=(1, 75, 50, 100)