File size: 6,003 Bytes
b30c649
 
 
a47c569
 
 
 
 
 
 
 
 
 
 
 
b30c649
a47c569
004ed06
 
a47c569
 
b30c649
004ed06
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a47c569
004ed06
a47c569
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
004ed06
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a47c569
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b30c649
004ed06
 
 
 
 
 
 
 
 
 
b30c649
004ed06
 
 
 
 
b30c649
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
004ed06
 
 
 
b30c649
 
 
004ed06
 
 
a47c569
b30c649
 
004ed06
 
 
b30c649
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
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
---
language:
- en
license: cc-by-4.0
task_categories:
- video-text-to-text
tags:
- gaze
- multimodal
- video-understanding
- streaming-video
- temporal-reasoning
- proactive-understanding
- egocentric-vision
- visual-question-answering
---

# StreamGaze Dataset

[Paper](https://huggingface.co/papers/2512.01707) | [Project Page](https://streamgaze.github.io/) | [Code](https://github.com/daeunni/StreamGaze)

**StreamGaze** is a comprehensive streaming video benchmark for evaluating MLLMs on gaze-based QA tasks across past, present, and future contexts.

## πŸ“ Dataset Structure

```
streamgaze/
β”œβ”€β”€ metadata/
β”‚   β”œβ”€β”€ egtea.csv              # EGTEA fixation metadata
β”‚   β”œβ”€β”€ egoexolearn.csv        # EgoExoLearn fixation metadata
β”‚   └── holoassist.csv         # HoloAssist fixation metadata
β”‚
β”œβ”€β”€ qa/
β”‚   β”œβ”€β”€ past_gaze_sequence_matching.json
β”‚   β”œβ”€β”€ past_non_fixated_object_identification.json
β”‚   β”œβ”€β”€ past_object_transition_prediction.json
β”‚   β”œβ”€β”€ past_scene_recall.json
β”‚   β”œβ”€β”€ present_future_action_prediction.json
β”‚   β”œβ”€β”€ present_object_attribute_recognition.json
β”‚   β”œβ”€β”€ present_object_identification_easy.json
β”‚   β”œβ”€β”€ present_object_identification_hard.json
β”‚   β”œβ”€β”€ proactive_gaze_triggered_alert.json
β”‚   └── proactive_object_appearance_alert.json
β”‚
└── videos/
    β”œβ”€β”€ videos_egtea_original.tar.gz         # EGTEA original videos
    β”œβ”€β”€ videos_egtea_viz.tar.gz              # EGTEA with gaze visualization
    β”œβ”€β”€ videos_egoexolearn_original.tar.gz   # EgoExoLearn original videos
    β”œβ”€β”€ videos_egoexolearn_viz.tar.gz        # EgoExoLearn with gaze visualization
    β”œβ”€β”€ videos_holoassist_original.tar.gz    # HoloAssist original videos
    └── videos_holoassist_viz.tar.gz         # HoloAssist with gaze visualization
```

## 🎯 Task Categories

### **Past (Historical Context)**
- **Gaze Sequence Matching**: Match gaze patterns to action sequences
- **Non-Fixated Object Identification**: Identify objects outside gaze
- **Object Transition Prediction**: Predict object state changes
- **Scene Recall**: Recall scene details from memory

### **Present (Current Context)**
- **Object Identification (Easy/Hard)**: Identify objects in/outside FOV
- **Object Attribute Recognition**: Recognize object attributes
- **Future Action Prediction**: Predict upcoming actions

### **Proactive (Future-Oriented)**
- **Gaze-Triggered Alert**: Alert based on gaze patterns
- **Object Appearance Alert**: Alert on object appearance

## πŸš€ Quick Start

### Data Preparation

Download our dataset from HuggingFace and extract videos. The dataset should be located as below (note: the dataset itself is in `danaleee/StreamGaze`):

```
StreamGaze/
β”œβ”€β”€ dataset/
β”‚   β”œβ”€β”€ videos/
β”‚   β”‚   β”œβ”€β”€ original_video/        # Original egocentric videos
β”‚   β”‚   └── gaze_viz_video/        # Videos with gaze overlay
β”‚   └── qa/
β”‚       β”œβ”€β”€ past_*.json             # Past task QA pairs
β”‚       β”œβ”€β”€ present_*.json          # Present task QA pairs
β”‚       └── proactive_*.json        # Proactive task QA pairs
```

#### Extract Videos

```bash
# Extract EGTEA videos
tar -xzf videos_egtea_original.tar.gz -C videos/egtea/original/
tar -xzf videos_egtea_viz.tar.gz -C videos/egtea/viz/

# Extract EgoExoLearn videos
tar -xzf videos_egoexolearn_original.tar.gz -C videos/egoexolearn/original/
tar -xzf videos_egoexolearn_viz.tar.gz -C videos/egoexolearn/viz/

# Extract HoloAssist videos
tar -xzf videos_holoassist_original.tar.gz -C videos/holoassist/original/
tar -xzf videos_holoassist_viz.tar.gz -C videos/holoassist/viz/
```

### Running Evaluation

Quick evaluation on existing models:

```bash
# Evaluate ViSpeak (without gaze visualization)
bash scripts/vispeak.sh

# Evaluate ViSpeak (with gaze visualization)
bash scripts/vispeak.sh --use_gaze_instruction

# Evaluate GPT-4o
bash scripts/gpt4o.sh --use_gaze_instruction

# Evaluate Qwen2.5-VL
bash scripts/qwen25vl.sh --use_gaze_instruction
```

Results will be automatically computed and saved to:

```
results/
β”œβ”€β”€ ModelName/
β”‚   β”œβ”€β”€ results/              # Without gaze visualization
β”‚   β”‚   β”œβ”€β”€ *_output.json
β”‚   β”‚   └── evaluation_summary.json
β”‚   └── results_viz/          # With gaze visualization
β”‚       β”œβ”€β”€ *_output.json
β”‚       └── evaluation_summary.json
```

## πŸ”‘ Metadata Format

Each metadata CSV contains:
- `video_source`: Video identifier
- `fixation_id`: Fixation segment ID
- `start_time_seconds` / `end_time_seconds`: Temporal boundaries
- `center_x` / `center_y`: Gaze center coordinates (normalized)
- `representative_object`: Primary object at gaze point
- `other_objects_in_cropped_area`: Objects within FOV
- `other_objects_outside_fov`: Objects outside FOV
- `scene_caption`: Scene description
- `action_caption`: Action description 

## πŸ“ QA Format

Each QA JSON file contains:
```json
  {
    "response_time": "[00:08 - 09:19]",
    "questions": [
      {
        "question": "Among {milk, spoon, pan, phone}, which did the user never gaze at?",
        "time_stamp": "03:14",
        "answer": "A",
        "options": [
          "A. milk",
          "B. spoon",
          "C. pan",
          "D. phone"
        ],
      }
    ],
    "video_path": "OP01-R03-BaconAndEggs.mp4"
  }
```

## πŸ“„ License

This dataset is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license.

See https://creativecommons.org/licenses/by/4.0/

## πŸ”— Links

- **Paper**: [https://huggingface.co/papers/2512.01707](https://huggingface.co/papers/2512.01707)
- **Evaluation code**: [https://github.com/daeunni/StreamGaze](https://github.com/daeunni/StreamGaze)
- **Project page**: [https://streamgaze.github.io/](https://streamgaze.github.io/)

## πŸ“§ Contact

For questions or issues, please contact: daeun@cs.unc.edu