mirror of
https://github.com/azaion/ai-training.git
synced 2026-04-22 07:06:36 +00:00
zxsanny
50 lines
4.0 KiB
Python
50 lines
4.0 KiB
Python
import re
|
|
|
|
import yaml
|
|
import constants
|
|
from azaion_api import Api, ApiCredentials
|
|
from inference.inference import Inference
|
|
from inference.onnx_engine import OnnxEngine
|
|
from inference.tensorrt_engine import TensorRTEngine
|
|
from utils import Dotdict
|
|
|
|
if __name__ == "__main__":
|
|
# Inference(OnnxEngine('azaion-2025-03-10.onnx', batch_size=4),
|
|
# confidence_threshold=0.5, iou_threshold=0.3).process('ForAI_test.mp4')
|
|
# detection for the first 200sec of video:
|
|
# onnxInference: 81 sec, 6.3Gb VRAM
|
|
# tensorrt: 54 sec, 3.7Gb VRAM
|
|
|
|
# Inference(TensorRTEngine('azaion-2025-03-10_int8.engine', batch_size=16),
|
|
# confidence_threshold=0.5, iou_threshold=0.3).process('ForAI_test.mp4')
|
|
# INT8 for 200sec: 54 sec 3.7Gb
|
|
|
|
# Inference(TensorRTEngine('azaion-2025-03-10_batch8.engine', batch_size=8),
|
|
# confidence_threshold=0.5, iou_threshold=0.3).process('ForAI_test.mp4')
|
|
|
|
with open(constants.CONFIG_FILE, "r") as f:
|
|
config_dict = yaml.safe_load(f)
|
|
d_config = Dotdict(config_dict)
|
|
cdn_c = Dotdict(d_config.cdn)
|
|
api_c = Dotdict(d_config.api)
|
|
api_client = Api(ApiCredentials(api_c.url, api_c.user, api_c.pw, api_c.folder))
|
|
|
|
tensor_model_bytes = api_client.load_resource(constants.AI_TENSOR_MODEL_FILE_BIG, constants.AI_TENSOR_MODEL_FILE_SMALL)
|
|
onxx_model_bytes = api_client.load_resource(constants.AI_ONNX_MODEL_FILE_BIG, constants.AI_ONNX_MODEL_FILE_SMALL)
|
|
|
|
input_string2 = "{'0': 'Armored-Vehicle', '1': 'Truck', '2': 'Vehicle', '3': 'Artillery', '4': 'Shadow', '5': 'Trenches', '6': 'Military-men', '7': 'Tyre-tracks', '8': 'Additional-armored-tank', '9': 'Smoke', '10': 'Plane', '11': 'Moto', '12': 'Camouflage-net', '13': 'Camouflage-branches', '14': 'Class-15', '15': 'Class-16', '16': 'Class-17', '17': 'Class-18', '18': 'Class-19', '19': 'Class-20'}"
|
|
result_dict2 = eval(input_string2)
|
|
|
|
try:
|
|
input_string3 = "{'0': 'Armored-Vehicle', '1': 'Truck', '2': 'Vehicle', '3': 'Artillery', '4': 'Shadow', '5': 'Trenches', '6': 'Military-men', '7': 'Tyre-tracks', '8': 'Additional-armored-tank', '9': 'Smoke', '10': 'Plane', '11': 'Moto', '12': 'Camouflage-net', '13': 'Camouflage-branches', '14': 'Class-15', '15': 'Class-16', '16': 'Class-17', '17': 'Class-18', '18': 'Class-19', '19': 'Class-20', '20': 'Armored-Vehicle(Wint)', '21': 'Truck(Wint)', '22': 'Vehicle(Wint)', '23': 'Artillery(Wint)', '24': 'Shadow(Wint)', '25': 'Trenches(Wint)', '26': 'Military-men(Wint)', '27': 'Tyre-tracks(Wint)', '28': 'Additional-armored-tank(Wint)', '29': 'Smoke(Wint)', '30': 'Plane(Wint)', '31': 'Moto(Wint)', '32': 'Camouflage-net(Wint)', '33': 'Camouflage-branches(Wint)', '34': 'Class-35', '35': 'Class-36', '36': 'Class-37', '37': 'Class-38', '38': 'Class-39', '39': 'Class-40', '40': 'Armored-Vehicle(Night)', '41': 'Truck(Night)', '42': 'Vehicle(Night)', '43': 'Artillery(Night)', '44': 'Shadow(Night)', '45': 'Trenches(Night)', '46': 'Military-men(Night)', '47': 'Tyre-tracks(Night)', '48': 'Additional-armored-tank(Night)', '49': 'Smoke(Night)', '50': 'Plane(Night)', '51': 'Moto(Night)', '52': 'Camouflage-net(Night)', '53': 'Camouflage-branches(Night)', '54': 'Class-55', '55': 'Class-56', '56': 'Class-57', '57': 'Class-58', '58': 'Class-59', '59': 'Class-60', '60': 'Class-61', '61': 'Class-62', '62': 'Class-63', '63': 'Class-64', '64': 'Class-65', '65': 'Class-66', '66': 'Class-67', '67': 'Class-68', '68': 'Class-69', '69': 'Class-70', '70': 'Class-71', '71': 'Class-72', '72': 'Class-73', '73': 'Class-74', '74': 'Class-75', '75': 'Class-76', '76': 'Class-77', '77': 'Class-78', '78': 'Class-79', '79': 'Class-80'}"
|
|
result_dict3 = eval(input_string3)
|
|
print(result_dict3)
|
|
except Exception as e:
|
|
print(e)
|
|
|
|
# Inference(OnnxEngine(onxx_model_bytes, batch_size=4),
|
|
# confidence_threshold=0.5, iou_threshold=0.3).process('tests/ForAI_test.mp4')
|
|
|
|
Inference(TensorRTEngine(tensor_model_bytes, batch_size=4),
|
|
confidence_threshold=0.5, iou_threshold=0.3).process('tests/ForAI_test.mp4')
|