File size: 3,572 Bytes
6ccbddf
 
6210d9e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fc1b8d0
 
 
 
 
 
 
 
 
 
6ccbddf
c6bf63d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
---
license: apache-2.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: video
    dtype: video
  - name: frames_response
    dtype: string
  - name: sliced_frames
    list: image
  splits:
  - name: train
    num_bytes: 1928709371
    num_examples: 500
  download_size: 1928580946
  dataset_size: 1928709371
task_categories:
- video-text-to-text
language:
- en
tags:
- Gym-Exercise
- Video-Analysis
pretty_name: ' Gym-Exercise'
size_categories:
- n<1K
---

# Gym-Exercise-Video-Analysis

**Gym-Exercise-Video-Analysis** is a specialized multimodal video understanding dataset comprising 500 annotated gym workout and exercise clips. It is designed for fine-tuning and evaluating Video-Language Models (Video-LLMs), visual fitness coaches, and temporal exercise analysis systems. Each entry pairs exercise videos and extracted frame sequences with in-depth textual descriptions, biomechanical observations, form evaluations, and routine tracking.

- **Curator:** [prithivMLmods](https://huggingface.co/prithivMLmods)
- **Total Samples:** 500 rows
- **Total Size:** ~1.93 GB
- **Format:** Parquet (`video`, `sliced_frames`, `frames_response`)
- **Modalities:** Image, Video, Text
- **Split:** Train (500 rows)

## Dataset Structure & Schema

Each record contains raw video data, sampled/sliced temporal frames, and comprehensive step-by-step descriptive analysis.

### Feature Fields

| Field | Type | Description |
| :--- | :--- | :--- |
| `video` | `Video` | Source video file of the exercise execution |
| `sliced_frames` | `Sequence[Image]` | List of sampled sequential video frames (typically 5 keyframes per clip) |
| `frames_response` | `string` | Detailed analysis describing the exercise movement, technique, body posture, and equipment used |

### Example Analysis Text

> *"The video captures an individual performing a seated workout routine... As you monitor your fitness routine, I can clearly see that the movement maintains steady tempo, targeted engagement of the upper body muscles, and controlled eccentric extension."*

## How to Use

### Loading with `datasets`

```python
from datasets import load_dataset

# Load dataset
dataset = load_dataset("prithivMLmods/Gym-Exercise-Video-Analysis", split="train")

# Access a single record
sample = dataset[0]
sliced_frames = sample["sliced_frames"]  # List of PIL Images
analysis = sample["frames_response"]     # Text analysis

print("Analysis preview:", analysis[:200])
print(f"Extracted keyframes: {len(sliced_frames)}")
```

### Video-LLM Fine-Tuning Format Example

Convert records into multi-image or video prompt conversations for models like Qwen2-VL, Video-LLaVA, or LLaVA-OneVision:

```python
def format_for_video_llm(example):
    return {
        "images": example["sliced_frames"],
        "prompt": "Analyze this gym exercise sequence. Identify the movement, assess form, and describe the physical execution in detail.",
        "response": example["frames_response"]
    }

formatted_sample = format_for_video_llm(dataset[0])
```

## Intended Uses

* **Video-LLM Alignment:** Instruction tuning multimodal models on multi-frame sequential reasoning and dense video captioning.
* **AI Fitness & Coaching Assistants:** Training automated gym form-checkers, exercise counters, and workout logging models.
* **Action & Movement Recognition:** Temporal motion understanding across diverse gym environments, lighting conditions, and workout equipment.

## License

This dataset is distributed under the **Apache-2.0 License**.