diff --git a/official/cv/Arcface/eval_ijbc.py b/official/cv/Arcface/eval_ijbc.py
index 26f5180b15892eedcfc758fc3780958f5176f10b..a0ef724d4ac08515f2b6319e00f360b563522b64 100644
--- a/official/cv/Arcface/eval_ijbc.py
+++ b/official/cv/Arcface/eval_ijbc.py
@@ -156,8 +156,8 @@ def divideIntoNstrand(listTemp, n):
 
 def read_template_media_list(path):
     ijb_meta = pd.read_csv(path, sep=' ', header=None).values
-    templates = ijb_meta[:, 1].astype(np.int)
-    media = ijb_meta[:, 2].astype(np.int)
+    templates = ijb_meta[:, 1].astype(np.int_)
+    media = ijb_meta[:, 2].astype(np.int_)
     return templates, media
 
 
@@ -166,9 +166,9 @@ def read_template_media_list(path):
 
 def read_template_pair_list(path):
     pairs = pd.read_csv(path, sep=' ', header=None).values
-    t1 = pairs[:, 0].astype(np.int)
-    t2 = pairs[:, 1].astype(np.int)
-    label = pairs[:, 2].astype(np.int)
+    t1 = pairs[:, 0].astype(np.int_)
+    t2 = pairs[:, 1].astype(np.int_)
+    label = pairs[:, 2].astype(np.int_)
     return t1, t2, label
 
 
diff --git a/official/cv/Arcface/eval_ijbc_onnx.py b/official/cv/Arcface/eval_ijbc_onnx.py
index 317c51f4b5680755e3fb5e146dd1ec07e45c3901..b6ba4fbb18126cb95f70080ada08596bd01d6625 100644
--- a/official/cv/Arcface/eval_ijbc_onnx.py
+++ b/official/cv/Arcface/eval_ijbc_onnx.py
@@ -148,16 +148,16 @@ def divideIntoNstrand(listTemp, n):
 
 def read_template_media_list(path):
     ijb_meta = pd.read_csv(path, sep=' ', header=None).values
-    templates = ijb_meta[:, 1].astype(np.int)
-    media = ijb_meta[:, 2].astype(np.int)
+    templates = ijb_meta[:, 1].astype(np.int_)
+    media = ijb_meta[:, 2].astype(np.int_)
 
     return templates, media
 
 def read_template_pair_list(path):
     pairs = pd.read_csv(path, sep=' ', header=None).values
-    t1 = pairs[:, 0].astype(np.int)
-    t2 = pairs[:, 1].astype(np.int)
-    label = pairs[:, 2].astype(np.int)
+    t1 = pairs[:, 0].astype(np.int_)
+    t2 = pairs[:, 1].astype(np.int_)
+    label = pairs[:, 2].astype(np.int_)
 
     return t1, t2, label
 
diff --git a/official/cv/CTPN/src/text_connector/connect_text_lines.py b/official/cv/CTPN/src/text_connector/connect_text_lines.py
index 171beca9a7856a5922011a44f050d2e4e3554403..87fa238921b20b980fff8db5ba4d07022f1416c8 100644
--- a/official/cv/CTPN/src/text_connector/connect_text_lines.py
+++ b/official/cv/CTPN/src/text_connector/connect_text_lines.py
@@ -52,7 +52,7 @@ def connect_text_lines(text_proposals, scores, size):
 
     text_lines = clip_boxes(text_lines, size)
 
-    text_recs = np.zeros((len(text_lines), 9), np.float)
+    text_recs = np.zeros((len(text_lines), 9), np.float_)
     index = 0
     for line in text_lines:
         xmin, ymin, xmax, ymax = line[0], line[1], line[2], line[3]
diff --git a/official/cv/CTPN/src/text_connector/detector.py b/official/cv/CTPN/src/text_connector/detector.py
index 7e5d724d440bc0cd8d6eab816bc053b51c30a8b6..707876a6773f837102b034cd02ffd6f0e26d8796 100644
--- a/official/cv/CTPN/src/text_connector/detector.py
+++ b/official/cv/CTPN/src/text_connector/detector.py
@@ -44,9 +44,9 @@ def filter_boxes(boxes):
     Returns:
         boxes(numpy.array): Text boxes after filter.
     """
-    heights = np.zeros((len(boxes), 1), np.float)
-    widths = np.zeros((len(boxes), 1), np.float)
-    scores = np.zeros((len(boxes), 1), np.float)
+    heights = np.zeros((len(boxes), 1), np.float_)
+    widths = np.zeros((len(boxes), 1), np.float_)
+    scores = np.zeros((len(boxes), 1), np.float_)
     index = 0
     for box in boxes:
         widths[index] = abs(box[2] - box[0])
diff --git a/official/cv/Efficientnet/efficientnet-b0/src/transform_utils.py b/official/cv/Efficientnet/efficientnet-b0/src/transform_utils.py
index 4e86cf9fbd7141941f019e3e733f6dcf57612295..ef3092cf9915826fa261a020492d07811b79ff78 100644
--- a/official/cv/Efficientnet/efficientnet-b0/src/transform_utils.py
+++ b/official/cv/Efficientnet/efficientnet-b0/src/transform_utils.py
@@ -421,7 +421,7 @@ def skew(img, v, **__):
         matrix.append([p1[0], p1[1], 1, 0, 0, 0, -p2[0] * p1[0], -p2[0] * p1[1]])
         matrix.append([0, 0, 0, p1[0], p1[1], 1, -p2[1] * p1[0], -p2[1] * p1[1]])
 
-    A = np.matrix(matrix, dtype=np.float)
+    A = np.matrix(matrix, dtype=np.float_)
     B = np.array(original_plane).reshape(8)
     perspective_skew_coefficients_matrix = np.dot(np.linalg.pinv(A), B)
     perspective_skew_coefficients_matrix = np.array(perspective_skew_coefficients_matrix).reshape(8)
diff --git a/official/cv/Efficientnet/efficientnet-b3/infer/util/classification_task_metric.py b/official/cv/Efficientnet/efficientnet-b3/infer/util/classification_task_metric.py
index 9e689cde780b1a8c7722b29656b0f73f7cdcaedd..ce60a7353792ddd6d92f16dd6d86bbd3bf75580b 100644
--- a/official/cv/Efficientnet/efficientnet-b3/infer/util/classification_task_metric.py
+++ b/official/cv/Efficientnet/efficientnet-b3/infer/util/classification_task_metric.py
@@ -55,7 +55,7 @@ def load_statistical_predict_result(filepath):
         data_vec = np.zeros((len(label_list)), dtype=np.float32)
         if n_label != 0:
             for ind, cls_ind in enumerate(label_list):
-                data_vec[ind] = np.int(cls_ind)
+                data_vec[ind] = np.int_(cls_ind)
     return data_vec, n_label
 
 
diff --git a/official/cv/FasterRCNN/src/detecteval.py b/official/cv/FasterRCNN/src/detecteval.py
index a6766af9712fc1e54b0fe6fadbee1b5b010d6635..63d9b21a33a7ddc38d5a35d4350311ec3f1bccc5 100644
--- a/official/cv/FasterRCNN/src/detecteval.py
+++ b/official/cv/FasterRCNN/src/detecteval.py
@@ -499,8 +499,8 @@ class DetectEval(COCOeval):
                     assert (tps.shape[0]) == 1
                     assert (fps.shape[0]) == 1
 
-                    tp_sum = np.cumsum(tps, axis=1).astype(dtype=np.float)
-                    fp_sum = np.cumsum(fps, axis=1).astype(dtype=np.float)
+                    tp_sum = np.cumsum(tps, axis=1).astype(dtype=np.float_)
+                    fp_sum = np.cumsum(fps, axis=1).astype(dtype=np.float_)
                     ids = catIds[k0]
                     label = labels[ids]
 
diff --git a/official/cv/Inception/inceptionv4/infer/sdk/classification_task_metric.py b/official/cv/Inception/inceptionv4/infer/sdk/classification_task_metric.py
index 09d0184a85bc4df6a82ff33c368bee7c05ab2ccb..4fdf1ec2b3c16a4129b52cc388fe7cadb1a34f2f 100644
--- a/official/cv/Inception/inceptionv4/infer/sdk/classification_task_metric.py
+++ b/official/cv/Inception/inceptionv4/infer/sdk/classification_task_metric.py
@@ -48,7 +48,7 @@ def load_statistical_predict_result(filepath):
         data_vec = np.zeros((n_label), dtype=np.float32)
         if n_label != 0:
             for ind, cls_ind in enumerate(temp):
-                data_vec[ind] = np.int(cls_ind)
+                data_vec[ind] = np.int_(cls_ind)
     return data_vec, n_label
 
 
diff --git a/official/cv/OCRNet/src/basedataset.py b/official/cv/OCRNet/src/basedataset.py
index 104ee830f131389cf35e7e5b033cbd1df131b8ec..cb676d79336365f1dfe97f1c906640925deee33e 100644
--- a/official/cv/OCRNet/src/basedataset.py
+++ b/official/cv/OCRNet/src/basedataset.py
@@ -89,14 +89,14 @@ class BaseDataset:
 
     def multi_scale_aug(self, image, label=None, rand_scale=1, rand_crop=True):
         """Augment feature into different scales."""
-        long_size = np.int(self.base_size * rand_scale + 0.5)
+        long_size = np.int_(self.base_size * rand_scale + 0.5)
         h, w, _ = image.shape
         if h > w:
             new_h = long_size
-            new_w = np.int(w * long_size / h + 0.5)
+            new_w = np.int_(w * long_size / h + 0.5)
         else:
             new_w = long_size
-            new_h = np.int(h * long_size / w + 0.5)
+            new_h = np.int_(h * long_size / w + 0.5)
 
         image = cv2.resize(image, (new_w, new_h), interpolation=cv2.INTER_LINEAR)
         # image = cv2.resize(image, (new_w, new_h), interpolation=cv2.INTER_NEAREST)
@@ -156,8 +156,8 @@ class BaseDataset:
         batch, _, ori_height, ori_width = image.shape
         assert batch == 1, "only supporting batchsize 1."
         image = image.asnumpy()[0].transpose((1, 2, 0)).copy()
-        stride_h = np.int(self.crop_size[0] * 2.0 / 3.0)
-        stride_w = np.int(self.crop_size[1] * 2.0 / 3.0)
+        stride_h = np.int_(self.crop_size[0] * 2.0 / 3.0)
+        stride_w = np.int_(self.crop_size[1] * 2.0 / 3.0)
 
         final_pred = Tensor(np.zeros([1, self.num_classes, ori_height, ori_width]), dtype=dtype.float32)
         padvalue = -1.0 * np.array(self.mean) / np.array(self.std)
@@ -178,10 +178,10 @@ class BaseDataset:
                     new_img = self.pad_image(new_img, height, width,
                                              self.crop_size, padvalue)
                 new_h, new_w = new_img.shape[:-1]
-                rows = np.int(np.ceil(1.0 * (new_h -
-                                             self.crop_size[0]) / stride_h)) + 1
-                cols = np.int(np.ceil(1.0 * (new_w -
-                                             self.crop_size[1]) / stride_w)) + 1
+                rows = np.int_(np.ceil(1.0 * (new_h -
+                                              self.crop_size[0]) / stride_h)) + 1
+                cols = np.int_(np.ceil(1.0 * (new_w -
+                                              self.crop_size[1]) / stride_w)) + 1
                 preds = Tensor(np.zeros([1, self.num_classes, new_h, new_w]), dtype=dtype.float32)
                 count = Tensor(np.zeros([1, 1, new_h, new_w]), dtype=dtype.float32)
 
diff --git a/official/cv/OCRNet/src/cityscapes.py b/official/cv/OCRNet/src/cityscapes.py
index 3dbe7a51944fad9b74a09a53a9a6f25a5ddb92a6..18b3a335d3f71c697103130e34ee4a92208fceb3 100644
--- a/official/cv/OCRNet/src/cityscapes.py
+++ b/official/cv/OCRNet/src/cityscapes.py
@@ -118,8 +118,8 @@ class Cityscapes(BaseDataset):
         batch, _, ori_height, ori_width = image.shape
         assert batch == 1, "only supporting batchsize 1."
         image = image.asnumpy()[0].transpose((1, 2, 0)).copy()
-        stride_h = np.int(self.crop_size[0] * 1.0)
-        stride_w = np.int(self.crop_size[1] * 1.0)
+        stride_h = np.int_(self.crop_size[0] * 1.0)
+        stride_w = np.int_(self.crop_size[1] * 1.0)
 
         final_pred = Tensor(np.zeros([1, self.num_classes, ori_height, ori_width]), dtype=dtype.float32)
         for scale in scales:
@@ -137,10 +137,10 @@ class Cityscapes(BaseDataset):
                 preds = preds[:, :, 0:height, 0:width]
             else:
                 new_h, new_w = new_img.shape[:-1]
-                rows = np.int(np.ceil(1.0 * (new_h -
-                                             self.crop_size[0]) / stride_h)) + 1
-                cols = np.int(np.ceil(1.0 * (new_w -
-                                             self.crop_size[1]) / stride_w)) + 1
+                rows = np.int_(np.ceil(1.0 * (new_h -
+                                              self.crop_size[0]) / stride_h)) + 1
+                cols = np.int_(np.ceil(1.0 * (new_w -
+                                              self.crop_size[1]) / stride_w)) + 1
                 preds = np.zeros([1, self.num_classes, new_h, new_w]).astype(np.float32)
 
                 count = np.zeros([1, 1, new_h, new_w]).astype(np.float32)
diff --git a/official/cv/OCRNet/src/seg_hrnet.py b/official/cv/OCRNet/src/seg_hrnet.py
index 49d3740abfc3d438b5cc02a9ca93224984b8decc..3fbcc063270ef2fc27fac9cd3f6e7e751cdaed9f 100644
--- a/official/cv/OCRNet/src/seg_hrnet.py
+++ b/official/cv/OCRNet/src/seg_hrnet.py
@@ -340,7 +340,7 @@ class HighResolutionNet(nn.Cell):
         self.stage4, pre_stage_channels = self._make_stage(
             self.stage4_cfg, num_channels, multi_scale_output=True)
 
-        last_inp_channels = np.int(np.sum(pre_stage_channels))
+        last_inp_channels = np.int_(np.sum(pre_stage_channels))
 
 
         self.last_layer = nn.SequentialCell([
diff --git a/official/cv/OCRNet/src/seg_hrnet_ocr.py b/official/cv/OCRNet/src/seg_hrnet_ocr.py
index 6cbd664a71daab6758c18d4cb5e2a5fd7c43859a..275e623f98c8617360c73df338562a880ae25228 100644
--- a/official/cv/OCRNet/src/seg_hrnet_ocr.py
+++ b/official/cv/OCRNet/src/seg_hrnet_ocr.py
@@ -565,7 +565,7 @@ class HighResolutionNet(nn.Cell):
         self.stage4, pre_stage_channels = self._make_stage(
             self.stage4_cfg, num_channels, multi_scale_output=True)
 
-        last_inp_channels = np.int(np.sum(pre_stage_channels))
+        last_inp_channels = np.int_(np.sum(pre_stage_channels))
         ocr_mid_channels = config.model.ocr.mid_channels
         ocr_key_channels = config.model.ocr.key_channels
 
diff --git a/official/cv/ResNet/infer/ResNet152/sdk/classification_task_metric.py b/official/cv/ResNet/infer/ResNet152/sdk/classification_task_metric.py
index 4c82151c1ad0b0fd8322b8e704e670a871963c27..a830f7d2cfd87c1a4ba6d8e6d79cd514ce49880b 100644
--- a/official/cv/ResNet/infer/ResNet152/sdk/classification_task_metric.py
+++ b/official/cv/ResNet/infer/ResNet152/sdk/classification_task_metric.py
@@ -69,7 +69,7 @@ def load_statistical_predict_result(filepath):
         data_vec = np.zeros((len(temp)), dtype=np.float32)
         if n_label != 0:
             for ind, cls_ind in enumerate(temp):
-                data_vec[ind] = np.int(cls_ind)
+                data_vec[ind] = np.int_(cls_ind)
     return data_vec, n_label
 
 
diff --git a/official/cv/RetinaFace_ResNet50/eval.py b/official/cv/RetinaFace_ResNet50/eval.py
index 6e79c5f29e3e6ce5c79e9996f35ce97a97a28905..671447b416df84bf450e5b8fb5094dba10cfc730 100644
--- a/official/cv/RetinaFace_ResNet50/eval.py
+++ b/official/cv/RetinaFace_ResNet50/eval.py
@@ -148,8 +148,8 @@ class DetectionEngine:
         keep = self._nms(dets, self.nms_thresh)
         dets = dets[keep, :]
 
-        dets[:, 2:4] = (dets[:, 2:4].astype(np.int) - dets[:, 0:2].astype(np.int)).astype(np.float) # int
-        dets[:, 0:4] = dets[:, 0:4].astype(np.int).astype(np.float)                                 # int
+        dets[:, 2:4] = (dets[:, 2:4].astype(np.int_) - dets[:, 0:2].astype(np.int_)).astype(np.float_) # int
+        dets[:, 0:4] = dets[:, 0:4].astype(np.int_).astype(np.float_)                                 # int
 
 
         # add to result
@@ -157,7 +157,7 @@ class DetectionEngine:
         if event_name not in self.results.keys():
             self.results[event_name] = {}
         self.results[event_name][img_name[:-4]] = {'img_path': image_path,
-                                                   'bboxes': dets[:, :5].astype(np.float).tolist()}
+                                                   'bboxes': dets[:, :5].astype(np.float_).tolist()}
 
     def _get_gt_boxes(self):
         from scipy.io import loadmat
@@ -182,7 +182,7 @@ class DetectionEngine:
 
         for event in self.results:
             for name in self.results[event].keys():
-                bbox = np.array(self.results[event][name]['bboxes']).astype(np.float)
+                bbox = np.array(self.results[event][name]['bboxes']).astype(np.float_)
                 if bbox.shape[0] <= 0:
                     continue
                 max_score = max(max_score, np.max(bbox[:, -1]))
@@ -191,7 +191,7 @@ class DetectionEngine:
         length = max_score - min_score
         for event in self.results:
             for name in self.results[event].keys():
-                bbox = np.array(self.results[event][name]['bboxes']).astype(np.float)
+                bbox = np.array(self.results[event][name]['bboxes']).astype(np.float_)
                 if bbox.shape[0] <= 0:
                     continue
                 bbox[:, -1] -= min_score
@@ -227,7 +227,7 @@ class DetectionEngine:
 
 
 
-        image_pr = np.zeros((section_num, 2), dtype=np.float)
+        image_pr = np.zeros((section_num, 2), dtype=np.float_)
         for section in range(section_num):
             _thresh = 1 - (section + 1)/section_num
             over_score_index = np.where(predict[:, 4] >= _thresh)[0]
@@ -254,7 +254,7 @@ class DetectionEngine:
         for _set in range(len(sets)):
             gt_list = set_gts[_set]
             count_gt = 0
-            pr_curve = np.zeros((section_num, 2), dtype=np.float)
+            pr_curve = np.zeros((section_num, 2), dtype=np.float_)
             for i, _ in enumerate(event_list):
                 event = str(event_list[i][0][0])
                 image_list = file_list[i][0]
@@ -263,7 +263,7 @@ class DetectionEngine:
                 event_gt_box_list = facebox_list[i][0]
 
                 for j, _ in enumerate(image_list):
-                    predict = np.array(event_predict_dict[str(image_list[j][0][0])]['bboxes']).astype(np.float)
+                    predict = np.array(event_predict_dict[str(image_list[j][0][0])]['bboxes']).astype(np.float_)
                     gt_boxes = event_gt_box_list[j][0].astype('float')
                     keep_index = event_gt_index_list[j][0]
                     count_gt += len(keep_index)
diff --git a/official/cv/Unet/postprocess.py b/official/cv/Unet/postprocess.py
index 67def617c84f07193552924990106ec26b61cdf8..8fd6d2da3c8f111a4501ce8e9b038a09f8ba9302 100644
--- a/official/cv/Unet/postprocess.py
+++ b/official/cv/Unet/postprocess.py
@@ -39,7 +39,7 @@ if __name__ == '__main__':
             mask = cv2.imread(os.path.join(config.data_path, f, "mask.png"), cv2.IMREAD_GRAYSCALE)
             mask = cv2.resize(mask, img_size)
             mask = mask.astype(np.float32) / 255
-            mask = (mask > 0.5).astype(np.int)
+            mask = (mask > 0.5).astype(np.int_)
             mask = (np.arange(2) == mask[..., None]).astype(int)
             mask = mask.transpose(2, 0, 1).astype(np.float32)
             label = mask.reshape(1, 2, 96, 96)
diff --git a/official/cv/Unet/src/data_loader.py b/official/cv/Unet/src/data_loader.py
index f2fc3eb15e81f670d24a3e4006ba16b63d98398c..e90863f6622a4866c393464a046a1d7b193ef98b 100644
--- a/official/cv/Unet/src/data_loader.py
+++ b/official/cv/Unet/src/data_loader.py
@@ -32,7 +32,7 @@ def _load_multipage_tiff(path):
 def _get_val_train_indices(length, fold, ratio=0.8):
     assert 0 < ratio <= 1, "Train/total data ratio must be in range (0.0, 1.0]"
     np.random.seed(0)
-    indices = np.arange(0, length, 1, dtype=np.int)
+    indices = np.arange(0, length, 1, dtype=np.int_)
     np.random.shuffle(indices)
 
     if fold is not None:
@@ -49,7 +49,7 @@ def _get_val_train_indices(length, fold, ratio=0.8):
 def data_post_process(img, mask):
 
     img = np.expand_dims(img, axis=0)
-    mask = (mask > 0.5).astype(np.int)
+    mask = (mask > 0.5).astype(np.int_)
     mask = (np.arange(mask.max() + 1) == mask[..., None]).astype(int)
     mask = mask.transpose(2, 0, 1).astype(np.float32)
     return img, mask
@@ -238,9 +238,9 @@ def preprocess_img_mask(img, mask, num_classes, img_size, augment=False, eval_re
     img = img.transpose(2, 0, 1)
     if num_classes == 2:
         mask = mask.astype(np.float32) / mask.max()
-        mask = (mask > 0.5).astype(np.int)
+        mask = (mask > 0.5).astype(np.int_)
     else:
-        mask = mask.astype(np.int)
+        mask = mask.astype(np.int_)
     mask = (np.arange(num_classes) == mask[..., None]).astype(int)
     mask = mask.transpose(2, 0, 1).astype(np.float32)
     return img, mask
diff --git a/official/cv/VIT/src/autoaugment.py b/official/cv/VIT/src/autoaugment.py
index 737e1945a8551c3cd99f8886dce7f20ee02555e1..fc5426f4f29152527f94982bda53df44ea11265c 100644
--- a/official/cv/VIT/src/autoaugment.py
+++ b/official/cv/VIT/src/autoaugment.py
@@ -207,7 +207,7 @@ class SubPolicy():
             "translateY": np.linspace(0, 150 / 331, 10),
             "rotate": np.linspace(0, 30, 10),
             "color": np.linspace(0.0, 0.9, 10),
-            "posterize": np.round(np.linspace(8, 4, 10), 0).astype(np.int),
+            "posterize": np.round(np.linspace(8, 4, 10), 0).astype(np.int_),
             "solarize": np.linspace(256, 0, 10),
             "contrast": np.linspace(0.0, 0.9, 10),
             "sharpness": np.linspace(0.0, 0.9, 10),
diff --git a/official/cv/YOLOv4/src/coco_visual.py b/official/cv/YOLOv4/src/coco_visual.py
index ea9459295092e0128612dde1c6398a0ef588bf33..a15a70ee2ff2739a1287c8426d141eb5d0217904 100644
--- a/official/cv/YOLOv4/src/coco_visual.py
+++ b/official/cv/YOLOv4/src/coco_visual.py
@@ -551,8 +551,8 @@ class DetectEval(COCOeval):
                     assert (tps.shape[0]) == 1
                     assert (fps.shape[0]) == 1
 
-                    tp_sum = np.cumsum(tps, axis=1).astype(dtype=np.float)
-                    fp_sum = np.cumsum(fps, axis=1).astype(dtype=np.float)
+                    tp_sum = np.cumsum(tps, axis=1).astype(dtype=np.float_)
+                    fp_sum = np.cumsum(fps, axis=1).astype(dtype=np.float_)
                     ids = catIds[k0]
                     label = labels[ids]