个人笔记记录v8 推理

使用v8_v2进行推理时。在ultralytics/data/loaders.py文件第270行添加。
#sxxpredict
# mx = getMulti2Array_255_apart(path)
# im0 = np.concatenate((im0, mx), axis=2)
再搭配

model = YOLO("./runs/4090/train" + str(index) + "/weights/best.pt")
path= "../../v8_multi/datasets/v3/devide302_ecc_vi_devide_cls1/test/images")
model.predict(source=path,ch=8,save=True, save_txt=True,save_conf=True)

ultralytics/engine/predictor.py 239行3改为8

#sxxpredict
self.model.warmup(imgsz=(1 if self.model.pt or self.model.triton else self.dataset.bs, 8, *self.imgsz))