Pikachu commited on
Commit
0eaf780
·
verified ·
1 Parent(s): e593dde

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +33 -21
README.md CHANGED
@@ -1,36 +1,48 @@
1
- ---
2
- license: apache-2.0
3
  ---
4
 
5
- *This dataset is still under construction as I'm re-recording the dataset but with a specific and consistent angle I can consistently recreate*
6
 
7
- # Minecraft Fly-over dataset 🎈
8
- This is a Minecraft dataset that contains a video of ~4.5 hours of flying over minecraft terrain.
9
 
10
  ---
11
 
12
- ## Specifications
 
 
 
 
13
 
14
  - **Resolution**: 2048x2048
15
- - **Frame Rate**: 30fps
16
  - **Field of View (FOV)**: 85 degrees
17
  - **Minecraft Version**: 1.20.4
18
- - **Day/Night Cycle**: Enabled (Day Cycle only, no weather cycle)
19
- - **Camera Orientation**: Almost facing the sunrise.
20
- - **Biomes Covered**: The video features an extensive range of biomes
21
 
22
- ### The footage was captured continuously without any interruptions
23
- You are free to use this footage for
24
 
25
- - **Machine Learning**: Train models using this dataset for image or video generation, etc
26
- - **Cinematics**: Use the footage for timelapses or anything related.
27
- - And more!
28
 
29
- ### Additional Notes
30
 
31
- - **No Modifications**: The Minecraft world was generated using default settings, with no modifications or custom biomes added.
32
- - **Data warning**: Given the high resolution and frame rate, the dataset is quite big (~34gb!)
33
- - **Collissions**: Because I wanted to capture quite low, I had to set up a few command blocks that removes every block in a certain radius of me. You might see blocks dissapearing and the camera go through a mountain.
34
- - **Mods**: Used Fabric with Distant Horizons and Sodium.
 
 
 
 
 
 
 
 
 
 
35
 
36
- When doing ML or AI applications, please let me know and keep me up to date!
 
1
+ Sure! Here's a polished version of the description:
2
+
3
  ---
4
 
5
+ # Minecraft Fly-over Dataset 🎈
6
 
7
+ *⚠️ Please note: This dataset is a work in progress and may be updated at any time. ⚠️*
 
8
 
9
  ---
10
 
11
+ ## Overview
12
+
13
+ This dataset offers a stunning, continuous 4.5-hour video capturing a fly-over of diverse Minecraft terrain.
14
+
15
+ ### Key Features
16
 
17
  - **Resolution**: 2048x2048
18
+ - **Frame Rate**: 30 fps
19
  - **Field of View (FOV)**: 85 degrees
20
  - **Minecraft Version**: 1.20.4
21
+ - **Day/Night Cycle**: Enabled (Day and night only, no weather effects)
22
+ - **Camera Orientation**: Focused slightly towards the sunrise
23
+ - **Biomes Covered**: The video spans a wide variety of biomes
24
 
25
+ ### Footage Details
 
26
 
27
+ - **Uninterrupted Recording**: The entire video was captured in one continuous take.
28
+ - **Unmodified World**: The Minecraft world was generated using default settings with no customizations or modded biomes.
29
+ - **Camera Positioning**: Command blocks were used to remove obstacles, which may cause visible blocks to disappear or the camera to "clip" through terrain.
30
 
31
+ ### Usage
32
 
33
+ This dataset can be used for:
34
+
35
+ - **Machine Learning**: Train models for image or video generation, terrain analysis, etc.
36
+ - **Cinematic Projects**: Use the footage for timelapses, animations, or other creative endeavors.
37
+ - **And more**: Use it for backgrounds, anything that you can do with it is allowed.
38
+
39
+ ### Additional Information
40
+
41
+ - **Size**: Due to its high resolution and video length, the dataset is quite large (~34GB).
42
+ - **Stuttering**: Slight stutters may occur because of the constant loading and unloading of chunks during the fly-over. These can be minimized or removed with code.
43
+ - **Limited Angle**: The video captures the world from a single, consistent angle.
44
+ - **Biome Coverage**: While a big chunk of biomes are included, not all biomes may be in the dataset, particularly rarer ones.
45
+
46
+ ---
47
 
48
+ Feel free to use this footage in your projects, and if you apply it to machine learning or AI work, I'd love to hear about your results!