Datasets:
Tasks:
Video Classification
Modalities:
Text
Formats:
webdataset
Languages:
English
Size:
1K - 10K
License:
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,118 @@
|
|
1 |
-
---
|
2 |
-
license: apache-2.0
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
task_categories:
|
4 |
+
- video-classification
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
tags:
|
8 |
+
- Sports
|
9 |
+
- Cricket
|
10 |
+
- Classification
|
11 |
+
size_categories:
|
12 |
+
- n<1K
|
13 |
+
---
|
14 |
+
# CricketShotClassification Dataset
|
15 |
+
|
16 |
+
## Dataset Description
|
17 |
+
|
18 |
+
This dataset is designed for video classification of cricket shots. It contains labeled videos of ten different cricket shots, making it suitable for training and evaluating machine learning models for cricket action recognition.
|
19 |
+
|
20 |
+
|
21 |
+
## Dataset Structure
|
22 |
+
|
23 |
+
The dataset contains videos of ten cricket shots:
|
24 |
+
|
25 |
+
| Shot Name | Label | Class ID |
|
26 |
+
|---------------|------------|-----------|
|
27 |
+
| Cover Drive | cover | 0 |
|
28 |
+
| Defense Shot | defense | 1 |
|
29 |
+
| Flick Shot | flick | 2 |
|
30 |
+
| Hook Shot | hook | 3 |
|
31 |
+
| Late Cut | late_cut | 4 |
|
32 |
+
| Lofted Drive | lofted | 5 |
|
33 |
+
| Pull Shot | pull | 6 |
|
34 |
+
| Square Cut | square_cut | 7 |
|
35 |
+
| Straight Drive| straight | 8 |
|
36 |
+
| Sweep Shot | sweep | 9 |
|
37 |
+
|
38 |
+
### Data Splits
|
39 |
+
- Train: 1,250 samples (125 per class)
|
40 |
+
- Validation: 250 samples (25 per class)
|
41 |
+
- Test: 250 samples (25 per class)
|
42 |
+
|
43 |
+
### Directory Structure
|
44 |
+
```
|
45 |
+
dataset/
|
46 |
+
βββ train/
|
47 |
+
β βββ cover/
|
48 |
+
β βββ defense/
|
49 |
+
β βββ flick/
|
50 |
+
β βββ hook/
|
51 |
+
β βββ late_cut/
|
52 |
+
β βββ lofted/
|
53 |
+
β βββ pull/
|
54 |
+
β βββ square_cut/
|
55 |
+
β βββ straight/
|
56 |
+
β βββ sweep/
|
57 |
+
βββ val/
|
58 |
+
β βββ cover/
|
59 |
+
β βββ defense/
|
60 |
+
β βββ flick/
|
61 |
+
β βββ hook/
|
62 |
+
β βββ late_cut/
|
63 |
+
β βββ lofted/
|
64 |
+
β βββ pull/
|
65 |
+
β βββ square_cut/
|
66 |
+
β βββ straight/
|
67 |
+
β βββ sweep/
|
68 |
+
βββ test/
|
69 |
+
βββ cover/
|
70 |
+
βββ defense/
|
71 |
+
βββ flick/
|
72 |
+
βββ hook/
|
73 |
+
βββ late_cut/
|
74 |
+
βββ lofted/
|
75 |
+
βββ pull/
|
76 |
+
βββ square_cut/
|
77 |
+
βββ straight/
|
78 |
+
βββ sweep/
|
79 |
+
```
|
80 |
+
|
81 |
+
## Usage
|
82 |
+
|
83 |
+
```python
|
84 |
+
from datasets import load_dataset
|
85 |
+
|
86 |
+
# Load the dataset from the hub
|
87 |
+
dataset = load_dataset("rokmr/cricketshot")
|
88 |
+
|
89 |
+
# Access different splits
|
90 |
+
train_data = dataset['train']
|
91 |
+
val_data = dataset['validation']
|
92 |
+
test_data = dataset['test']
|
93 |
+
|
94 |
+
```
|
95 |
+
|
96 |
+
## Limitations and Bias
|
97 |
+
|
98 |
+
- Limited sample size per class
|
99 |
+
- May not cover all possible variations of each shot
|
100 |
+
- Lighting and camera angle variations may affect model performance
|
101 |
+
|
102 |
+
## Citation
|
103 |
+
If you use this dataset, please cite:
|
104 |
+
```
|
105 |
+
@misc{cricketshot,
|
106 |
+
author = {Rohit Kumar},
|
107 |
+
title = {Cricket Shot Classification Dataset},
|
108 |
+
year = {2025},
|
109 |
+
publisher = {Hugging Face}
|
110 |
+
}
|
111 |
+
```
|
112 |
+
|
113 |
+
## Acknowledgements
|
114 |
+
The dataset structure and initial inspiration was drawn from the [CricShot10 dataset](https://github.com/ascuet/CricShot10) We thank the authors for making their work publicly available.
|
115 |
+
|
116 |
+
|
117 |
+
## Dataset Card Contact
|
118 |
+
For questions or concerns about the dataset, please contact the maintainers through the Hugging Face community platform.
|