dima806 commited on
Commit
38b2c96
·
verified ·
1 Parent(s): 4db43cf

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +23 -21
README.md CHANGED
@@ -8,31 +8,33 @@ Return hand gesture based on image.
8
 
9
  See https://www.kaggle.com/code/dima806/hand-gestures-image-detection-vit for more details.
10
 
 
 
11
  ```
12
  Classification report:
13
 
14
  precision recall f1-score support
15
 
16
- call 0.8748 0.9097 0.8919 2957
17
- dislike 0.9679 0.9577 0.9628 2956
18
- fist 0.9628 0.9811 0.9718 2956
19
- four 0.8241 0.8068 0.8154 2956
20
- like 0.9039 0.8559 0.8793 2957
21
- mute 0.9608 0.9956 0.9779 2956
22
- ok 0.8929 0.9053 0.8990 2956
23
- one 0.9266 0.9175 0.9220 2957
24
- palm 0.9531 0.9202 0.9363 2957
25
- peace 0.8576 0.8620 0.8598 2956
26
- peace_inverted 0.9672 0.9472 0.9571 2957
27
- rock 0.8962 0.8471 0.8710 2957
28
- stop 0.9092 0.9591 0.9335 2956
29
- stop_inverted 0.9661 0.9151 0.9399 2956
30
- three 0.7879 0.8258 0.8064 2956
31
- three2 0.9204 0.9188 0.9196 2957
32
- two_up 0.9014 0.9642 0.9317 2957
33
- two_up_inverted 0.9649 0.9381 0.9513 2956
34
 
35
- accuracy 0.9126 53216
36
- macro avg 0.9132 0.9126 0.9126 53216
37
- weighted avg 0.9132 0.9126 0.9126 53216
38
  ```
 
8
 
9
  See https://www.kaggle.com/code/dima806/hand-gestures-image-detection-vit for more details.
10
 
11
+ ![image/png](https://cdn-uploads.huggingface.co/production/uploads/6449300e3adf50d864095b90/HAOiyjaETTnpeq78UEfXp.png)
12
+
13
  ```
14
  Classification report:
15
 
16
  precision recall f1-score support
17
 
18
+ call 0.8781 0.9189 0.8980 11826
19
+ dislike 0.9623 0.9669 0.9646 11826
20
+ fist 0.9678 0.9849 0.9763 11825
21
+ four 0.8296 0.8163 0.8229 11825
22
+ like 0.9200 0.8546 0.8861 11826
23
+ mute 0.9615 0.9945 0.9777 11825
24
+ ok 0.9020 0.9112 0.9066 11826
25
+ one 0.9304 0.9226 0.9265 11825
26
+ palm 0.9505 0.9271 0.9387 11825
27
+ peace 0.8491 0.8683 0.8586 11826
28
+ peace_inverted 0.9690 0.9428 0.9557 11826
29
+ rock 0.8951 0.8490 0.8714 11825
30
+ stop 0.9111 0.9614 0.9356 11826
31
+ stop_inverted 0.9717 0.9181 0.9441 11826
32
+ three 0.7985 0.8279 0.8129 11826
33
+ three2 0.9291 0.9249 0.9270 11825
34
+ two_up 0.9100 0.9608 0.9347 11826
35
+ two_up_inverted 0.9618 0.9373 0.9494 11826
36
 
37
+ accuracy 0.9160 212861
38
+ macro avg 0.9165 0.9160 0.9159 212861
39
+ weighted avg 0.9165 0.9160 0.9159 212861
40
  ```