hi .sry to bother u again~ about emotion_classifier

[shillyshallysxy]

Sorry for my bad English, I hope I could covey to you my thoughts accurately. That’s because I preprocessed the input picture and recognized them simultaneously, and I summed the probability of output without normalization. If there were three inputs after preprocessing, the sum of the probabilities will be three. You could just divide the result with this number and you will get the output you want. ------------------ 原始邮件 ------------------ 发件人: "Cyrus"<[email protected]> 发送时间: 2019年4月26日(星期五) 下午3:45 收件人: "shillyshallysxy/emotion_classifier"<[email protected]>; 抄送: "Subscribed"<[email protected]>; 主题: [shillyshallysxy/emotion_classifier] hi .sry to bother u again~ (#2) sry .i didnt install chinese input-method on my ubuntu OS so that i write down my question in ENGLISH. i run your code successfully on ubuntu os, and i tested ur model. the outcomes are as follows {'angry': 1.680927, 'disgust:': 0.024156317, 'fear': 0.38532764, 'happy': 0.37007406, 'sad': 0.30601713, 'surprise': 0.08675081, 'neutral': 0.146747} Emotion : angry {'angry': 0.2911712, 'disgust:': 0.044982724, 'fear': 0.24095652, 'happy': 1.2809961, 'sad': 0.17451024, 'surprise': 0.19956648, 'neutral': 0.7678167} Emotion : happy {'angry': 0.85339236, 'disgust:': 0.18872109, 'fear': 0.7736851, 'happy': 0.21916914, 'sad': 0.16908921, 'surprise': 0.25328648, 'neutral': 0.5426566} Emotion : angry im so confused that why the sum of these probabilities is not 1? im looking forward to ur early reply .thx~ &mdash; You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[O-Givenchy]

Sorry for my bad English, I hope I could covey to you my thoughts accurately. That’s because I preprocessed the input picture and recognized them simultaneously, and I summed the probability of output without normalization. If there were three inputs after preprocessing, the sum of the probabilities will be three. You could just divide the result with this number and you will get the output you want. ------------------ 原始邮件 ------------------
发件人: "Cyrus"[email protected]
发送时间: 2019年4月26日(星期五) 下午3:45
收件人: "shillyshallysxy/emotion_classifier"[email protected];
抄送: "Subscribed"[email protected];
主题: [shillyshallysxy/emotion_classifier] hi .sry to bother u again~ (#2)

sry .i didnt install chinese input-method on my ubuntu OS so that i write down my question in ENGLISH.

i run your code successfully on ubuntu os, and i tested ur model. the outcomes are as follows

{'angry': 1.680927, 'disgust:': 0.024156317, 'fear': 0.38532764, 'happy': 0.37007406, 'sad': 0.30601713, 'surprise': 0.08675081, 'neutral': 0.146747}
Emotion : angry
{'angry': 0.2911712, 'disgust:': 0.044982724, 'fear': 0.24095652, 'happy': 1.2809961, 'sad': 0.17451024, 'surprise': 0.19956648, 'neutral': 0.7678167}
Emotion : happy
{'angry': 0.85339236, 'disgust:': 0.18872109, 'fear': 0.7736851, 'happy': 0.21916914, 'sad': 0.16908921, 'surprise': 0.25328648, 'neutral': 0.5426566}
Emotion : angry

im so confused that why the sum of these probabilities is not 1?

im looking forward to ur early reply .thx~

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.

您好博主~

我刚刚把系统换回来了，可以打中文字了哈哈~

我算了一下。确实是您说的这样。
谢谢博主~:)