Update README.md
Browse files
README.md
CHANGED
@@ -44,8 +44,92 @@ dataset_info:
|
|
44 |
sequence: string
|
45 |
splits:
|
46 |
- name: train
|
47 |
-
num_bytes: 26654549442
|
48 |
num_examples: 195000
|
49 |
download_size: 12966047661
|
50 |
-
dataset_size: 26654549442
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
44 |
sequence: string
|
45 |
splits:
|
46 |
- name: train
|
47 |
+
num_bytes: 26654549442
|
48 |
num_examples: 195000
|
49 |
download_size: 12966047661
|
50 |
+
dataset_size: 26654549442
|
51 |
+
task_categories:
|
52 |
+
- image-to-image
|
53 |
+
- object-detection
|
54 |
+
tags:
|
55 |
+
- biology
|
56 |
+
- neuroscience
|
57 |
+
- fmri
|
58 |
+
size_categories:
|
59 |
+
- 100K<n<1M
|
60 |
---
|
61 |
+
|
62 |
+
# NSD-Flat
|
63 |
+
[[`GitHub`]](https://github.com/clane9/NSD-Flat) [[🤗 `Hugging Face Hub`]](https://huggingface.co/datasets/clane9/NSD-Flat)
|
64 |
+
|
65 |
+
A Hugging Face dataset of pre-processed brain activity flat maps from the [Natural Scenes Dataset](https://naturalscenesdataset.org/), constrained to a visual cortex region of interest and rendered as PNG images.
|
66 |
+
|
67 |
+
## Load the dataset
|
68 |
+
|
69 |
+
Load the dataset from [Hugging Face Hub](https://huggingface.co/datasets/clane9/NSD-Flat)
|
70 |
+
|
71 |
+
```python
|
72 |
+
from datasets import load_dataset
|
73 |
+
|
74 |
+
dataset = load_dataset("clane9/NSD-Flat", split="train")
|
75 |
+
```
|
76 |
+
|
77 |
+
## Building the dataset
|
78 |
+
|
79 |
+
### 1. Download source data
|
80 |
+
|
81 |
+
Run [`download_data.sh`](download_data.sh) to download the required source data:
|
82 |
+
|
83 |
+
- NSD stimuli images and presentation info
|
84 |
+
- COCO annotations
|
85 |
+
- NSD beta activity maps in fsaverge surface space
|
86 |
+
|
87 |
+
```bash
|
88 |
+
bash download_data.sh
|
89 |
+
```
|
90 |
+
|
91 |
+
### 2. Convert the COCO annotations
|
92 |
+
|
93 |
+
Run [`convert_nsd_annotations.py`](convert_nsd_annotations.py) to crop and reorganize the COCO annotations for NSD.
|
94 |
+
|
95 |
+
```bash
|
96 |
+
python convert_nsd_annotations.py
|
97 |
+
```
|
98 |
+
|
99 |
+
### 3. Generate the dataset
|
100 |
+
|
101 |
+
Run [`generate_dataset.py`](generate_dataset.py) to generate the huggingface dataset in Arrow format.
|
102 |
+
|
103 |
+
```bash
|
104 |
+
python generate_dataset.py --img_size 256 --workers 8
|
105 |
+
```
|
106 |
+
|
107 |
+
## Citation
|
108 |
+
|
109 |
+
If you find this dataset useful, please consider citing:
|
110 |
+
|
111 |
+
```
|
112 |
+
@article{allen2022massive,
|
113 |
+
title = {A massive 7T fMRI dataset to bridge cognitive neuroscience and artificial intelligence},
|
114 |
+
author = {Allen, Emily J and St-Yves, Ghislain and Wu, Yihan and Breedlove, Jesse L and Prince, Jacob S and Dowdle, Logan T and Nau, Matthias and Caron, Brad and Pestilli, Franco and Charest, Ian and others},
|
115 |
+
journal = {Nature neuroscience},
|
116 |
+
volume = {25},
|
117 |
+
number = {1},
|
118 |
+
pages = {116--126},
|
119 |
+
year = {2022},
|
120 |
+
publisher = {Nature Publishing Group US New York}
|
121 |
+
}
|
122 |
+
```
|
123 |
+
|
124 |
+
```
|
125 |
+
@misc{lane2023nsdflat,
|
126 |
+
author = {Connor Lane},
|
127 |
+
title = {NSD-Flat: Pre-processed brain activity flat maps from the Natural Scenes Dataset},
|
128 |
+
howpublished = {\url{https://huggingface.co/datasets/clane9/NSD-Flat}},
|
129 |
+
year = {2023},
|
130 |
+
}
|
131 |
+
```
|
132 |
+
|
133 |
+
## License
|
134 |
+
|
135 |
+
Usage of this dataset constitutes agreement to the [NSD Terms and Conditions](https://cvnlab.slite.page/p/IB6BSeW_7o/Terms-and-Conditions).
|