From fe0a154fe360b20cd72ea548921fa76bba1896b3 Mon Sep 17 00:00:00 2001 From: jiangxt <1579525634@qq.com> Date: Mon, 31 Jul 2023 10:42:49 +0800 Subject: [PATCH] =?UTF-8?q?7=E6=9C=8831=E6=97=A5=E6=9B=B4=E6=96=B0PP=5FTSM?= =?UTF-8?q?v2=E6=A8=A1=E5=9E=8B=E6=8E=A8=E7=90=86=E8=BE=93=E5=87=BA?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- tool/PP_TSMv2_infer.py | 107 ++++++++++++++++------------------------- 1 file changed, 41 insertions(+), 66 deletions(-) diff --git a/tool/PP_TSMv2_infer.py b/tool/PP_TSMv2_infer.py index e1f198a..401794a 100644 --- a/tool/PP_TSMv2_infer.py +++ b/tool/PP_TSMv2_infer.py @@ -1,29 +1,22 @@ -from export_model import trim_config,get_input_spec -from predict import parse_file_paths import os import os.path as osp -import paddle -from paddlevideo.utils import get_config -from paddlevideo.modeling.builder import build_model -from paddle.jit import to_static +from paddlevideo.utils.config import get_config from paddle.inference import Config, create_predictor from utils import build_inference_helper -import time import warnings warnings.filterwarnings("ignore") -class PP_TSMv2(object): +class PP_TSMv2_predict(object): """PP-TSMv2模型中常用的参数初始化""" - def __init__(self,use_gpu=True,batch_size=1,ir_optim=True,\ - disable_glog=False,save_name=None,enable_mklddn=False,\ + def __init__(self,use_gpu=True,ir_optim=True, + disable_glog=False,save_name=None,enable_mklddn=False, precision="fp32",gpu_mem=8000,cpu_threads=None): self.use_gpu = use_gpu #是否使用GPU self.cpu_threads = cpu_threads #cpu线程数 - self.batch_size = batch_size self.ir_optim = ir_optim #是否开启IR优化 self.disable_glog = disable_glog self.gpu_mem = gpu_mem #GPU存储大小 @@ -31,6 +24,25 @@ class PP_TSMv2(object): self.precision = precision #mfldnn精度 self.save_name = save_name #转化推理模型存放名称 + def parse_file_paths(self,input_path: str) -> list: + + """ + 获取模型输入数据 + input_path:模型的输入文件 + """ + if osp.isfile(input_path): + files = [ + input_path, + ] + else: + files = os.listdir(input_path) + files = [ + file for file in files + if (file.endswith(".avi") or file.endswith(".mp4")) + ] + files = [osp.join(input_path, file) for file in files] + return files + def create_paddle_predictor(self,model_f,pretr_p,cfg): """ @@ -66,47 +78,6 @@ class PP_TSMv2(object): return config,predictor - def exportmodel(self,config,pretr_p,output_p): - - """ - 加载训练后的模型参数，生成可推理预测的模型 - pretr_p:训练后的参数存放文件 - output_p:转化为可推理模型的存放路径 - """ - - cfg, model_name = trim_config(get_config(config, overrides=None, show=False)) - - print(f"Building model({model_name})...") - - #创建推理模型 - model = build_model(cfg.MODEL) - assert osp.isfile( - pretr_p - ), f"pretrained params ({pretr_p} is not a file path.)" - - if not os.path.isdir(output_p): - os.makedirs(output_p) - - print(f"Loading params from ({pretr_p})...") - - # 加载推理模型参数 - params = paddle.load(pretr_p) - model.set_dict(params) - - model.eval() - - for layer in model.sublayers(): - if hasattr(layer, "rep") and not getattr(layer, "is_repped"): - layer.rep() - - input_spec = get_input_spec(cfg.INFERENCE, model_name) - - #将模型转化为静态图以及模型的保存 - model = to_static(model, input_spec=input_spec) - paddle.jit.save(model,osp.join(output_p, model_name if self.save_name is None else self.save_name)) - print(f"model ({model_name}) has been already saved in ({output_p}).") - return model - def predict(self,config,input_f,batch_size,model_f,params_f): @@ -118,7 +89,7 @@ class PP_TSMv2(object): model_f:可推理模型存放的路径+配置文件 params_f:可推理模型的参数 """ - + result = {} cfg = get_config(config, overrides=None, show=False) model_name = cfg.model_name print(f"Inference model({model_name})...") @@ -139,7 +110,7 @@ class PP_TSMv2(object): for item in output_names: output_tensor_list.append(predictor.get_output_handle(item)) - files = parse_file_paths(input_f) + files = self.parse_file_paths(input_f)#input_path=input_f batch_num = batch_size for st_idx in range(0, len(files), batch_num): @@ -158,24 +129,28 @@ class PP_TSMv2(object): batched_outputs.append(output_tensor_list[j].copy_to_cpu()) #输出推理结果 - InferenceHelper.postprocess(batched_outputs,True) + res = InferenceHelper.postprocess(batched_outputs,False,True) + result["video_id"] = res[0]["video_id"] + result["topk_class"] = res[0]["topk_class"].tolist()[0] + result["topk_scores"] = res[0]["topk_scores"].tolist()[0] + + return result + def main(): - config='/home/xznsh/data/PaddleVideo/configs/recognition/pptsm/v2/pptsm_lcnet_k400_16frames_uniform.yaml' #配置文件地址 - input_file='/home/xznsh/data/PaddleVideo/data/dataset/video_seg_re_hand' #待推理数据集存放的地址 - pretrain_params='/home/xznsh/data/PaddleVideo/output/ppTSMv2/ppTSMv2_best.pdparams' #训练后模型参数文件存放 - output_path='/home/xznsh/data/PaddleVideo/inference/infer1' #推理模型存放地址 - model_file='/home/xznsh/data/PaddleVideo/inference/infer1/ppTSMv2.pdmodel' #推理模型存放地址 - params_file='/home/xznsh/data/PaddleVideo/inference/infer1/ppTSMv2.pdiparams' #推理模型参数存放地址 - batch_size= 1 - PP_TSMv2().exportmodel(config,pretrain_params,output_path) #输出推理模型 - time.sleep(2) - PP_TSMv2().predict(config,input_file,batch_size,model_file,params_file) #推理模型推理、预测 + config='D:/download/PaddleVideo1/configs/recognition/pptsm/v2/pptsm_lcnet_k400_16frames_uniform.yaml' #配置文件地址 + input_file='D:/download/PaddleVideo1/data/dataset/video_seg_no_hand/test02_84.avi' #待推理数据集存放的地址 + model_file='D:/download/PaddleVideo1/output/ppTSMv2.pdmodel' #推理模型存放地址 + params_file='D:/download/PaddleVideo1/output/ppTSMv2.pdiparams' #推理模型参数存放地址 + batch_size= 1 #输出推理模型 + t = PP_TSMv2_predict().predict(config,input_file,batch_size,model_file,params_file) #推理模型推理、预测 + print(t) + + - if __name__ == "__main__": main()