Merge branch 'main' of hf.co:datasets/alitourani/moviefeats into main
Browse files
README.md
CHANGED
|
@@ -1,155 +1,155 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: gpl-3.0
|
| 3 |
-
task_categories:
|
| 4 |
-
- feature-extraction
|
| 5 |
-
- image-classification
|
| 6 |
-
- video-classification
|
| 7 |
-
- image-feature-extraction
|
| 8 |
-
language:
|
| 9 |
-
- en
|
| 10 |
-
pretty_name: MoViFex_Dataset
|
| 11 |
-
size_categories:
|
| 12 |
-
-
|
| 13 |
-
---
|
| 14 |
-
|
| 15 |
-
# 🎬 MoViFex Dataset
|
| 16 |
-
|
| 17 |
-
The Movies Visual Features Extracted (MoViFex) dataset contains visual features obtained from a wide range of movies (full-length), their shots, and free trailers. It contains frame-level extracted visual features and aggregated version of them. **MoViFex** can be used in recommendation, information retrieval, classification, _etc_ tasks.
|
| 18 |
-
|
| 19 |
-
## 📃 Table of Content
|
| 20 |
-
|
| 21 |
-
- [How to Use](#usage)
|
| 22 |
-
- [Dataset Stats](#stats)
|
| 23 |
-
- [Files Structure](#structure)
|
| 24 |
-
|
| 25 |
-
## 🚀 How to Use? <a id="usage"></a>
|
| 26 |
-
|
| 27 |
-
### The Dataset Web-Page
|
| 28 |
-
|
| 29 |
-
Check the detailed information about the dataset in its web-page presented in the link in [https://recsys-lab.github.io/movifex_dataset/](https://recsys-lab.github.io/movifex_dataset/).
|
| 30 |
-
|
| 31 |
-
### The Designed Framework for Benchmarking
|
| 32 |
-
|
| 33 |
-
In order to use, exploit, and generate this dataset, a framework titled `MoViFex` is implemented. You can read more about it [on the GitHub repository](https://github.com/RecSys-lab/SceneSense).
|
| 34 |
-
|
| 35 |
-
## 📊 Dataset Stats <a id="stats"></a>
|
| 36 |
-
|
| 37 |
-
### General
|
| 38 |
-
|
| 39 |
-
| Aspect | Value |
|
| 40 |
-
| ----------------------------------------------- | --------- |
|
| 41 |
-
| **Total number of movies** | 274 |
|
| 42 |
-
| **Average frames extracted per movie** | 7,732 |
|
| 43 |
-
| **Total number of frames (or feature vectors)** | 2,118,647 |
|
| 44 |
-
|
| 45 |
-
### Hybrid (combined with **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)))
|
| 46 |
-
|
| 47 |
-
| Aspect | Value |
|
| 48 |
-
| ---------------------------------------- | --------- |
|
| 49 |
-
| **Accumulative number of genres:** | 723 |
|
| 50 |
-
| **Average movie ratings:** | 3.88/5 |
|
| 51 |
-
| **Total number of users:** | 158,146 |
|
| 52 |
-
| **Accumulative number of interactions:** | 2,869,024 |
|
| 53 |
-
|
| 54 |
-
### Required Capacity
|
| 55 |
-
|
| 56 |
-
| Data | Model | Total Files | Size on Disk |
|
| 57 |
-
| ---------------------- | ----- | ----------- | ------------- |
|
| 58 |
-
| Full Movies | incp3 | 84,872 | 35.8 GB |
|
| 59 |
-
| Full Movies | vgg19 | 84,872 | 46.1 GB |
|
| 60 |
-
| Movie Shots | incp3 | 16,713 | 7.01 GB |
|
| 61 |
-
| Movie Shots | vgg19 | 24,598 | 13.3 GB |
|
| 62 |
-
| Trailers | incp3 | 1,725 | 681 MB |
|
| 63 |
-
| Trailers | vgg19 | 1,725 | 885 MB |
|
| 64 |
-
| Aggregated Full Movies | incp3 | 84,872 | 10 MB |
|
| 65 |
-
| Aggregated Full Movies | vgg19 | 84,872 | 19 MB |
|
| 66 |
-
| Aggregated Movie Shots | incp3 | 16,713 | 10 MB |
|
| 67 |
-
| Aggregated Movie Shots | vgg19 | 24,598 | 19 MB |
|
| 68 |
-
| Aggregated Trailers | incp3 | 1,725 | 10 MB |
|
| 69 |
-
| Aggregated Trailers | vgg19 | 1,725 | 19 MB |
|
| 70 |
-
| **Total** | - | **214,505** | **~103.9 GB** |
|
| 71 |
-
|
| 72 |
-
## 🗄️ Files Structure <a id="structure"></a>
|
| 73 |
-
|
| 74 |
-
### Level I. Primary Categories
|
| 75 |
-
|
| 76 |
-
The dataset contains six main folders and a `stats.json` file. The `stats.json` file contains the meta-data for the sources. Folders **'full_movies'**, **'movie_shots'**, and **'movie_trailers'** keep the atomic visual features extracted from various sources, including `full_movies` for frame-level visual features extracted from full-length movie videos, `movie_shots` for the shot-level (_i.e.,_ important frames) visual features extracted from full-length movie videos, and `movie_trailers` for frame-level visual features extracted from movie trailers videos. Folders **'full_movies_agg'**, **'movie_shots_agg'**, and **'movie_trailers_agg'** keep the aggregated (non-atomic) versions of the described items.
|
| 77 |
-
|
| 78 |
-
### Level II. Visual Feature Extractors
|
| 79 |
-
|
| 80 |
-
Inside each of the mentioned folders, there are two folders titled `incp3` and `vgg19`, referring to the feature extractor used to generate the visual features, which are [Inception-v3 (GoogleNet)](https://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Szegedy_Rethinking_the_Inception_CVPR_2016_paper.html) and [VGG-19](https://doi.org/10.48550/arXiv.1409.1556), respectively.
|
| 81 |
-
|
| 82 |
-
### Level III. Contents (Movies & Trailers)
|
| 83 |
-
|
| 84 |
-
#### A: Atomic Features (folders full_movies, movie_shots, and movie_trailers)
|
| 85 |
-
|
| 86 |
-
Inside each feature extractor folder (_e.g.,_ `full_movies/incp3` or `movie_trailers/vgg19`) you can find a set of folders with unique title (_e.g.,_ `0000000778`) indicating the ID of the movie in **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)) dataset. Accordingly, you have access to the visual features extracted from the movie `0000000778`, using Inception-v3 and VGG-19 extractors, in full-length frame, full-length shot, and trailer levels.
|
| 87 |
-
|
| 88 |
-
#### B: Aggregated Features (folders full_movies_agg, movie_shots_agg, and movie_trailers_agg)
|
| 89 |
-
|
| 90 |
-
Inside each feature extractor folder (_e.g.,_ `full_movies_agg/incp3` or `movie_trailers_agg/vgg19`) you can find a set of `json` files with unique title (_e.g.,_ `0000000778.json`) indicating the ID of the movie in **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)) dataset. Accordingly, you have access to the aggregated visual features extracted from the movie `0000000778` (and available on the atomic features folders), using Inception-v3 and VGG-19 extractors, in full-length frame, full-length shot, and trailer levels.
|
| 91 |
-
|
| 92 |
-
### Level IV. Packets (Atomic Feature Folders Only)
|
| 93 |
-
|
| 94 |
-
To better organize visual features, each movie folder (_e.g.,_ `0000000778`) has a set of packets named as `packet0001.json` to `packet000N.json` saved as `json` files. Each packet contains a set of objects with `frameId` and `features` attributes, keeping the equivalent frame-ID and visual feature, respectively. In general, every **25** object (`frameId-features` pair) form a packet, except the last packet that can have less objects.
|
| 95 |
-
|
| 96 |
-
The described structure is presented below in brief:
|
| 97 |
-
|
| 98 |
-
```bash
|
| 99 |
-
> [full_movies] ## visual features of frame-level full-length movie videos
|
| 100 |
-
> [incp3] ## visual features extracted using Inception-v3
|
| 101 |
-
> [movie-1]
|
| 102 |
-
> [packet-1]
|
| 103 |
-
> [packet-2]
|
| 104 |
-
...
|
| 105 |
-
> [packet-m]
|
| 106 |
-
> [movie-2]
|
| 107 |
-
...
|
| 108 |
-
> [movie-n]
|
| 109 |
-
> [vgg19] ## visual features extracted using VGG-19
|
| 110 |
-
> [movie-1]
|
| 111 |
-
...
|
| 112 |
-
> [movie-n]
|
| 113 |
-
> [movie_shots] ## visual features of shot-level full-length movie videos
|
| 114 |
-
> [incp3]
|
| 115 |
-
> ...
|
| 116 |
-
> [vgg19]
|
| 117 |
-
> ...
|
| 118 |
-
> [movie_trailers] ## visual features of frame-level movie trailer videos
|
| 119 |
-
> [incp3]
|
| 120 |
-
> ...
|
| 121 |
-
> [vgg19]
|
| 122 |
-
> ...
|
| 123 |
-
> [full_movies_agg] ## aggregated visual features of frame-level full-length movie videos
|
| 124 |
-
> [incp3] ## aggregated visual features extracted using Inception-v3
|
| 125 |
-
> [movie-1-json]
|
| 126 |
-
> [movie-2]
|
| 127 |
-
...
|
| 128 |
-
> [movie-n]
|
| 129 |
-
> [vgg19] ## aggregated visual features extracted using VGG-19
|
| 130 |
-
> [movie-1]
|
| 131 |
-
...
|
| 132 |
-
> [movie-n]
|
| 133 |
-
> [movie_shots_agg] ## aggregated visual features of shot-level full-length movie videos
|
| 134 |
-
> [movie_trailers_agg] ## aggregated visual features of frame-level movie trailer videos
|
| 135 |
-
```
|
| 136 |
-
|
| 137 |
-
### `stats.json` File
|
| 138 |
-
|
| 139 |
-
The `stats.json` file placed in the root contains valuable information about the characteristics of each of the movies, fetched from **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)).
|
| 140 |
-
|
| 141 |
-
```json
|
| 142 |
-
[
|
| 143 |
-
{
|
| 144 |
-
"id": "0000000006",
|
| 145 |
-
"title": "Heat",
|
| 146 |
-
"year": 1995,
|
| 147 |
-
"genres": [
|
| 148 |
-
"Action",
|
| 149 |
-
"Crime",
|
| 150 |
-
"Thriller"
|
| 151 |
-
]
|
| 152 |
-
},
|
| 153 |
-
...
|
| 154 |
-
]
|
| 155 |
-
```
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: gpl-3.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- feature-extraction
|
| 5 |
+
- image-classification
|
| 6 |
+
- video-classification
|
| 7 |
+
- image-feature-extraction
|
| 8 |
+
language:
|
| 9 |
+
- en
|
| 10 |
+
pretty_name: MoViFex_Dataset
|
| 11 |
+
size_categories:
|
| 12 |
+
- n>100G
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
# 🎬 MoViFex Dataset
|
| 16 |
+
|
| 17 |
+
The Movies Visual Features Extracted (MoViFex) dataset contains visual features obtained from a wide range of movies (full-length), their shots, and free trailers. It contains frame-level extracted visual features and aggregated version of them. **MoViFex** can be used in recommendation, information retrieval, classification, _etc_ tasks.
|
| 18 |
+
|
| 19 |
+
## 📃 Table of Content
|
| 20 |
+
|
| 21 |
+
- [How to Use](#usage)
|
| 22 |
+
- [Dataset Stats](#stats)
|
| 23 |
+
- [Files Structure](#structure)
|
| 24 |
+
|
| 25 |
+
## 🚀 How to Use? <a id="usage"></a>
|
| 26 |
+
|
| 27 |
+
### The Dataset Web-Page
|
| 28 |
+
|
| 29 |
+
Check the detailed information about the dataset in its web-page presented in the link in [https://recsys-lab.github.io/movifex_dataset/](https://recsys-lab.github.io/movifex_dataset/).
|
| 30 |
+
|
| 31 |
+
### The Designed Framework for Benchmarking
|
| 32 |
+
|
| 33 |
+
In order to use, exploit, and generate this dataset, a framework titled `MoViFex` is implemented. You can read more about it [on the GitHub repository](https://github.com/RecSys-lab/SceneSense).
|
| 34 |
+
|
| 35 |
+
## 📊 Dataset Stats <a id="stats"></a>
|
| 36 |
+
|
| 37 |
+
### General
|
| 38 |
+
|
| 39 |
+
| Aspect | Value |
|
| 40 |
+
| ----------------------------------------------- | --------- |
|
| 41 |
+
| **Total number of movies** | 274 |
|
| 42 |
+
| **Average frames extracted per movie** | 7,732 |
|
| 43 |
+
| **Total number of frames (or feature vectors)** | 2,118,647 |
|
| 44 |
+
|
| 45 |
+
### Hybrid (combined with **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)))
|
| 46 |
+
|
| 47 |
+
| Aspect | Value |
|
| 48 |
+
| ---------------------------------------- | --------- |
|
| 49 |
+
| **Accumulative number of genres:** | 723 |
|
| 50 |
+
| **Average movie ratings:** | 3.88/5 |
|
| 51 |
+
| **Total number of users:** | 158,146 |
|
| 52 |
+
| **Accumulative number of interactions:** | 2,869,024 |
|
| 53 |
+
|
| 54 |
+
### Required Capacity
|
| 55 |
+
|
| 56 |
+
| Data | Model | Total Files | Size on Disk |
|
| 57 |
+
| ---------------------- | ----- | ----------- | ------------- |
|
| 58 |
+
| Full Movies | incp3 | 84,872 | 35.8 GB |
|
| 59 |
+
| Full Movies | vgg19 | 84,872 | 46.1 GB |
|
| 60 |
+
| Movie Shots | incp3 | 16,713 | 7.01 GB |
|
| 61 |
+
| Movie Shots | vgg19 | 24,598 | 13.3 GB |
|
| 62 |
+
| Trailers | incp3 | 1,725 | 681 MB |
|
| 63 |
+
| Trailers | vgg19 | 1,725 | 885 MB |
|
| 64 |
+
| Aggregated Full Movies | incp3 | 84,872 | 10 MB |
|
| 65 |
+
| Aggregated Full Movies | vgg19 | 84,872 | 19 MB |
|
| 66 |
+
| Aggregated Movie Shots | incp3 | 16,713 | 10 MB |
|
| 67 |
+
| Aggregated Movie Shots | vgg19 | 24,598 | 19 MB |
|
| 68 |
+
| Aggregated Trailers | incp3 | 1,725 | 10 MB |
|
| 69 |
+
| Aggregated Trailers | vgg19 | 1,725 | 19 MB |
|
| 70 |
+
| **Total** | - | **214,505** | **~103.9 GB** |
|
| 71 |
+
|
| 72 |
+
## 🗄️ Files Structure <a id="structure"></a>
|
| 73 |
+
|
| 74 |
+
### Level I. Primary Categories
|
| 75 |
+
|
| 76 |
+
The dataset contains six main folders and a `stats.json` file. The `stats.json` file contains the meta-data for the sources. Folders **'full_movies'**, **'movie_shots'**, and **'movie_trailers'** keep the atomic visual features extracted from various sources, including `full_movies` for frame-level visual features extracted from full-length movie videos, `movie_shots` for the shot-level (_i.e.,_ important frames) visual features extracted from full-length movie videos, and `movie_trailers` for frame-level visual features extracted from movie trailers videos. Folders **'full_movies_agg'**, **'movie_shots_agg'**, and **'movie_trailers_agg'** keep the aggregated (non-atomic) versions of the described items.
|
| 77 |
+
|
| 78 |
+
### Level II. Visual Feature Extractors
|
| 79 |
+
|
| 80 |
+
Inside each of the mentioned folders, there are two folders titled `incp3` and `vgg19`, referring to the feature extractor used to generate the visual features, which are [Inception-v3 (GoogleNet)](https://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Szegedy_Rethinking_the_Inception_CVPR_2016_paper.html) and [VGG-19](https://doi.org/10.48550/arXiv.1409.1556), respectively.
|
| 81 |
+
|
| 82 |
+
### Level III. Contents (Movies & Trailers)
|
| 83 |
+
|
| 84 |
+
#### A: Atomic Features (folders full_movies, movie_shots, and movie_trailers)
|
| 85 |
+
|
| 86 |
+
Inside each feature extractor folder (_e.g.,_ `full_movies/incp3` or `movie_trailers/vgg19`) you can find a set of folders with unique title (_e.g.,_ `0000000778`) indicating the ID of the movie in **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)) dataset. Accordingly, you have access to the visual features extracted from the movie `0000000778`, using Inception-v3 and VGG-19 extractors, in full-length frame, full-length shot, and trailer levels.
|
| 87 |
+
|
| 88 |
+
#### B: Aggregated Features (folders full_movies_agg, movie_shots_agg, and movie_trailers_agg)
|
| 89 |
+
|
| 90 |
+
Inside each feature extractor folder (_e.g.,_ `full_movies_agg/incp3` or `movie_trailers_agg/vgg19`) you can find a set of `json` files with unique title (_e.g.,_ `0000000778.json`) indicating the ID of the movie in **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)) dataset. Accordingly, you have access to the aggregated visual features extracted from the movie `0000000778` (and available on the atomic features folders), using Inception-v3 and VGG-19 extractors, in full-length frame, full-length shot, and trailer levels.
|
| 91 |
+
|
| 92 |
+
### Level IV. Packets (Atomic Feature Folders Only)
|
| 93 |
+
|
| 94 |
+
To better organize visual features, each movie folder (_e.g.,_ `0000000778`) has a set of packets named as `packet0001.json` to `packet000N.json` saved as `json` files. Each packet contains a set of objects with `frameId` and `features` attributes, keeping the equivalent frame-ID and visual feature, respectively. In general, every **25** object (`frameId-features` pair) form a packet, except the last packet that can have less objects.
|
| 95 |
+
|
| 96 |
+
The described structure is presented below in brief:
|
| 97 |
+
|
| 98 |
+
```bash
|
| 99 |
+
> [full_movies] ## visual features of frame-level full-length movie videos
|
| 100 |
+
> [incp3] ## visual features extracted using Inception-v3
|
| 101 |
+
> [movie-1]
|
| 102 |
+
> [packet-1]
|
| 103 |
+
> [packet-2]
|
| 104 |
+
...
|
| 105 |
+
> [packet-m]
|
| 106 |
+
> [movie-2]
|
| 107 |
+
...
|
| 108 |
+
> [movie-n]
|
| 109 |
+
> [vgg19] ## visual features extracted using VGG-19
|
| 110 |
+
> [movie-1]
|
| 111 |
+
...
|
| 112 |
+
> [movie-n]
|
| 113 |
+
> [movie_shots] ## visual features of shot-level full-length movie videos
|
| 114 |
+
> [incp3]
|
| 115 |
+
> ...
|
| 116 |
+
> [vgg19]
|
| 117 |
+
> ...
|
| 118 |
+
> [movie_trailers] ## visual features of frame-level movie trailer videos
|
| 119 |
+
> [incp3]
|
| 120 |
+
> ...
|
| 121 |
+
> [vgg19]
|
| 122 |
+
> ...
|
| 123 |
+
> [full_movies_agg] ## aggregated visual features of frame-level full-length movie videos
|
| 124 |
+
> [incp3] ## aggregated visual features extracted using Inception-v3
|
| 125 |
+
> [movie-1-json]
|
| 126 |
+
> [movie-2]
|
| 127 |
+
...
|
| 128 |
+
> [movie-n]
|
| 129 |
+
> [vgg19] ## aggregated visual features extracted using VGG-19
|
| 130 |
+
> [movie-1]
|
| 131 |
+
...
|
| 132 |
+
> [movie-n]
|
| 133 |
+
> [movie_shots_agg] ## aggregated visual features of shot-level full-length movie videos
|
| 134 |
+
> [movie_trailers_agg] ## aggregated visual features of frame-level movie trailer videos
|
| 135 |
+
```
|
| 136 |
+
|
| 137 |
+
### `stats.json` File
|
| 138 |
+
|
| 139 |
+
The `stats.json` file placed in the root contains valuable information about the characteristics of each of the movies, fetched from **MovieLenz 25M** ([link](https://grouplens.org/datasets/movielens/25m/)).
|
| 140 |
+
|
| 141 |
+
```json
|
| 142 |
+
[
|
| 143 |
+
{
|
| 144 |
+
"id": "0000000006",
|
| 145 |
+
"title": "Heat",
|
| 146 |
+
"year": 1995,
|
| 147 |
+
"genres": [
|
| 148 |
+
"Action",
|
| 149 |
+
"Crime",
|
| 150 |
+
"Thriller"
|
| 151 |
+
]
|
| 152 |
+
},
|
| 153 |
+
...
|
| 154 |
+
]
|
| 155 |
+
```
|