- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I am using customized YoloV3-tiny with 1 channel images (grayscale images) with different number of classes.
It works on darknet environment.
I noticed that yolo_v3.py and yolo_v3_tiny.py assumes 3channel image input and it might not work with models assumed grayscale image input.
Am I right?
Katsunori
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear Waragai, Katsunori
yolo_v3.py and yolo_v3_tiny.py are not part of the OpenVino installation. They are however part of the darknet repo. But I actually read through the code as well as the header comments in yolo_v3_tiny.py and I am not seeing why it shouldn't handle grayscale. C in this case would be "1" instead of "3". Doing a quick google search doesn't reveal any recent issues with grayscale either.
:param inputs: a 4-D tensor of size [batch_size, height, width, channels].
Dimension batch_size may be undefined. The channel order is RGB.
:param num_classes: number of predicted classes.
:param is_training: whether is training or not.
:param data_format: data format NCHW or NHWC.
:param reuse: whether or not the network and its variables should be reused.
:return:
"""
Hope it helps,
Thanks,
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you!
I succeeded in converting customized YoloV3-tiny with 1 channel images (grayscale images) with different number of classes.
https://software.intel.com/en-us/forums/computer-vision/topic/821024#comment-1945820
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Now I add --gray option for Inference script .
Katsunori
----
$ diff --cont object_detection_demo_yolov3_async_org.py object_detection_demo_yolov3_async.py
*** object_detection_demo_yolov3_async_org.py 2019-09-27 20:00:40.000000000 +0900
--- object_detection_demo_yolov3_async.py 2019-09-27 20:04:43.000000000 +0900
***************
*** 55,60 ****
--- 55,62 ----
action="store_true")
args.add_argument("-r", "--raw_output_message", help="Optional. Output inference results raw values showing",
default=False, action="store_true")
+ args.add_argument("-g", "--gray", help="Optional. Use grayscaled model",
+ default=False, action="store_true")
return parser
***************
*** 160,165 ****
--- 162,173 ----
return 0
return area_of_overlap / area_of_union
+ def one_channel_frame(frame):
+ if len(frame.shape) == 3:
+ frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+ h, w = frame.shape[:2]
+ frame = frame.reshape((h, w, 1))
+ return frame
def main():
args = build_argparser().parse_args()
***************
*** 218,223 ****
--- 226,233 ----
# Number of frames in picture is 1 and this will be read in cycle. Sync mode is default value for this case
if number_input_frames != 1:
ret, frame = cap.read()
+ if args.gray:
+ frame = one_channel_frame(frame)
else:
is_async_mode = False
wait_key_code = 0
***************
*** 240,247 ****
--- 250,261 ----
# in the regular mode, we capture frame to the CURRENT infer request
if is_async_mode:
ret, next_frame = cap.read()
+ if args.gray:
+ next_frame = one_channel_frame(next_frame)
else:
ret, frame = cap.read()
+ if args.gray:
+ frame = one_channel_frame(frame)
if not ret:
break
***************
*** 254,259 ****
--- 268,274 ----
in_frame = cv2.resize(frame, (w, h))
# resize input_frame to network size
+ in_frame = in_frame.reshape((h, w, c))
in_frame = in_frame.transpose((2, 0, 1)) # Change data layout from HWC to CHW
in_frame = in_frame.reshape((n, c, h, w))
***************
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear Waragai, Katsunori,
Thank you for sharing with the OpenVino community. Is your issue fixed now ?
Thanks,
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yes fixed.
![](/skins/images/C4F7392A5C68ACA60290C9576F107044/responsive_peak/images/icon_anonymous_message.png)
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page