TianzhongSong / Person-Segmentation-Keras

Person segmentation with Keras (SegNet, Unet, etc.)
Apache License 2.0
160 stars 34 forks source link

eval_segmentation.py #11

Open buaa-luzhi opened 5 years ago

buaa-luzhi commented 5 years ago

@TianzhongSong,您好! gt_img[:] += (gt[:] == 1).astype('uint8') pt_img[:] += (pt[:] == c).astype('uint8')

gt_img[:] shape=(1,50176, 2) gt[:] shape=(50176,)

pt_img[:] shape=(1,50176, 2) pt[:] shape=(50176,) 这应该也是个问题吧

python1806huqiugen commented 5 years ago

您好 我也遇到同样的问题,请问您解决了吗? File "eval_segmentation.py", line 56, in gt_img[:] += (gt[:] == 1).astype('uint8') ValueError: operands could not be broadcast together with shapes (1,50176,11) (50176,) (1,50176,11)

zkh418277390 commented 5 years ago

在 gt_img [:] + =(gt [:] == 1)中的文件“ eval_segmentation.py”(第56行).astype('uint8') ValueError 文件中 ,您好我也遇到同样的问题, 请问您解决了吗?:操作数不能与形状(1,50176,11)(50176,)(1,50176,11)一起广播 请问你解决了吗?

Chancetc commented 4 years ago

shape不一致,改成 gt_img = np.zeros_like(gt) pt_img = np.zeros_like(gt)

zkh418277390 commented 4 years ago

谢谢您,代码已经调通。

------------------ 原始邮件 ------------------ 发件人: "Chancetc"<notifications@github.com>; 发送时间: 2019年12月30日(星期一) 下午3:57 收件人: "TianzhongSong/Person-Segmentation-Keras"<Person-Segmentation-Keras@noreply.github.com>; 抄送: "418277390"<418277390@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [TianzhongSong/Person-Segmentation-Keras] eval_segmentation.py (#11)

shape不一致,改成 gt_img = np.zeros_like(gt) pt_img = np.zeros_like(gt)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

WuGe0424 commented 4 years ago

请问各位的代码跑出来是label标记的人形状么?我的预测完以后为什么都是椭圆状的,丝毫没有类似的人label形状

minushuang commented 2 years ago

请问各位的代码跑出来是label标记的人形状么?我的预测完以后为什么都是椭圆状的,丝毫没有类似的人label形状

是的,我的也是这样,请问后来解决了吗?