Closed hua812586 closed 1 year ago
Haven't tested it, but something like this should work:
import pandas as pd
from mmwave.model import LstmModel
def load_sample(path):
df = pd.read_csv(path)
num_of_frames = df.iloc[-1]['frame'] + 1
sample = [[] for _ in range(num_of_frames)]
for _, row in df.iterrows():
if row['x'] == 'None':
obj = 5*[0.]
else:
obj = [
float(row['x'])/65535.,
float(row['y'])/65535.,
float(row['range_idx'])/65535.,
float(row['peak_value'])/65535.,
float(row['doppler_idx'])/65535.
]
sample[row['frame']].append(obj)
return sample
model = LstmModel()
model.load() # loads model from `mmwave/.lstm_model`
model.predict([load_sample(<path>)])
Wow, thank you very much for your help. It can be used.
Hello @vilari-mickopf. Thanks for your great repository ‘mmwave-gesture-recognition’. Howover,I have a question that I would like to consult with you?
Now I can use the 1642 board to actually test gestures and accurately recognize them. But now I want to use the CSV file in the data for offline testing (that is, write a separate predicted file(.py), pass the CSV data in during runtime, and directly print out the results), but I don't know how to use the existing code to achieve it? Can you help me take a look?
Thanks for your help in advance. I wish to hear from you as soon as possible.