aidlab/sora_python
7
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

merge

Browse Source
This commit is contained in:
pangkaicheng
2024-03-28 10:44:58 +08:00
parent 8d34a4ed7f
commit 4e832bbadb
2 changed files with 19 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
app/service/outfit_matcher/outfit_evaluator.py
View File

@@ -379,7 +379,6 @@ class OutfitMaterTypeAware(OutfitMatcher):
Returns:
scores: List of float
"""
<<<<<<< HEAD
outfit_images, outfit_categories = self.preprocess(outfits, features)
scores = []
for images, categories in zip(outfit_images, outfit_categories):
@@ -400,28 +399,3 @@ class OutfitMaterTypeAware(OutfitMatcher):
scores = np.stack(scores, axis=0)
return scores.flatten()
=======
image, category, mask = self.preprocess(outfits)
client = httpclient.InferenceServerClient(url=f"{OM_TRITON_IP}:{OM_TRITON_PORT}")
# 输入集
inputs = [
httpclient.InferInput("input__0", image.shape, datatype="FP32"),
httpclient.InferInput("input__1", category.shape, datatype="INT16"),
httpclient.InferInput("input__2", mask.shape, datatype="FP32"),
]
inputs[0].set_data_from_numpy(image.astype(np.float32), binary_data=True)
inputs[1].set_data_from_numpy(category.astype(np.int16), binary_data=True)
inputs[2].set_data_from_numpy(mask.astype(np.float32), binary_data=True)
# 输出集
outputs = [
httpclient.InferRequestedOutput("output__0", binary_data=True),
httpclient.InferRequestedOutput("output__1", binary_data=True)
]
results = client.infer(model_name="outfit_matcher_type_aware", inputs=inputs, outputs=outputs)
# 推理
# 取结果
scores = torch.from_numpy(results.as_numpy("output__0")) # Shape (N, 1)
features = torch.from_numpy(results.as_numpy("output__1")) # Shape (N, 64)
return scores, features
>>>>>>> 1f23781b16e59bfbcbbb4d252e6a61685267e6c7

39
app/service/outfit_matcher/service.py
View File

@@ -18,38 +18,43 @@ if __name__ == '__main__':
all_items = param["query"] + param["database"]
unextracted_item = []
prepared_feature = {}
# 拿到所有需要提取特征的图片
for item in all_items:
if f'{item["item_name"]}.npy' not in os.listdir("feature"):
unextracted_item.append(item)
if len(unextracted_item) > 0:
# 通过backbone模型提取图片特征
extracted_features = backbone_service.get_result(unextracted_item)
for i, item in enumerate(unextracted_item):
np.save(f'feature/{item["item_name"]}.npy', extracted_features[i])
for item in all_items:
if item["item_name"] not in prepared_feature.keys():
prepared_feature[item["item_name"]] = np.load(f'feature/{item["item_name"]}.npy')
for item in tqdm(param["query"] * 10):
outfits = fashion_dataset.generate_outfit(item, param["topk"], param["max_outfits"])
<<<<<<< HEAD
scores = service.get_result(outfits, prepared_feature)
=======
scores, features = service.get_result(outfits)
# save features
# 链接milvus
# TODO
np.save(f'feature/{item["item_name"]}.npy', extracted_features[i])
# 存入数据库
# 关闭链接
>>>>>>> 1f23781b16e59bfbcbbb4d252e6a61685267e6c7
# print(scores)
# print(len(scores))
# TODO 读取本次任务需要的图片特征
for item in all_items:
if item["item_name"] not in prepared_feature.keys():
prepared_feature[item["item_name"]] = np.load(f'feature/{item["item_name"]}.npy')
# 开始服装搭配任务
for item in tqdm(param["query"] * 10):
# 根据一定规则生成outfit
outfits = fashion_dataset.generate_outfit(item, param["topk"], param["max_outfits"])
# 根据模型对生成的outfit打分
scores = service.get_result(outfits, prepared_feature)
# 对评分排序，拿到最好的topk个outfit输出
sorted_indices = np.argsort(scores)[:param["topk"]] # type-aware
outfits = [outfits[i] for i in sorted_indices] # 最好的五个
# 结果可视化
# service.visualize(outfits, scores, param["topk"], best=True,
# output_path=os.path.join(r"D:\PhD_Study\MIXI\mitu\image\123",
# f"{item['item_name']}_best_{param['topk']}.png"))
# service.visualize(outfits, scores, param["topk"], best=False,
# output_path=os.path.join(r"D:\PhD_Study\MIXI\mitu\image\123",
# f"{item['item_name']}_worst_{param['topk']}.png"))
sorted_indices = np.argsort(scores)[:param["topk"]] # type-aware
outfits = [outfits[i] for i in sorted_indices] # 最好的五个