{
    "name": "mtg_jamendo_moodtheme",
    "type": "multi-label classifier",
    "link": "https://essentia.upf.edu/models/classification-heads/mtg_jamendo_moodtheme/mtg_jamendo_moodtheme-discogs-effnet-1.pb",
    "version": "1",
    "description": "classification of music by mood and theme from discogs-effnet embeddings",
    "author": "Pablo Alonso",
    "email": "pablo.alonso@upf.edu",
    "release_date": "2022-11-20",
    "framework": "tensorflow",
    "framework_version": "2.8.0",
    "classes": [
        "action",
        "adventure",
        "advertising",
        "background",
        "ballad",
        "calm",
        "children",
        "christmas",
        "commercial",
        "cool",
        "corporate",
        "dark",
        "deep",
        "documentary",
        "drama",
        "dramatic",
        "dream",
        "emotional",
        "energetic",
        "epic",
        "fast",
        "film",
        "fun",
        "funny",
        "game",
        "groovy",
        "happy",
        "heavy",
        "holiday",
        "hopeful",
        "inspiring",
        "love",
        "meditative",
        "melancholic",
        "melodic",
        "motivational",
        "movie",
        "nature",
        "party",
        "positive",
        "powerful",
        "relaxing",
        "retro",
        "romantic",
        "sad",
        "sexy",
        "slow",
        "soft",
        "soundscape",
        "space",
        "sport",
        "summer",
        "trailer",
        "travel",
        "upbeat",
        "uplifting"
    ],
    "model_types": [
        "frozen_model"
    ],
    "dataset": {
        "name": "MTG Jamendo Dataset (mood and theme subset)",
        "size": "18,486 full tracks",
        "metrics": {
            "test PR-AUC": 0.14,
            "test ROC-AUC": 0.76
        }
    },
    "schema": {
        "inputs": [
            {
                "name": "model/Placeholder",
                "type": "float",
                "shape": [
                    1280
                ]
            }
        ],
        "outputs": [
            {
                "name": "model/Sigmoid",
                "type": "float",
                "shape": [
                    56
                ],
                "op": "Sigmoid",
                "output_purpose": "predictions"
            },
            {
                "name": "model/dense_1/BiasAdd",
                "type": "float",
                "shape": [
                    56
                ],
                "op": "fully connected",
                "description": "logits",
                "output_purpose": ""
            },
            {
                "name": "model/dense/BiasAdd",
                "type": "float",
                "shape": [
                    512
                ],
                "op": "fully connected",
                "description": "penultimate layer",
                "output_purpose": ""
            }
        ]
    },
    "citation": "@inproceedings{alonso2022music,\n  title={Music Representation Learning Based on Editorial Metadata from Discogs},\n  author={Alonso-Jim{\\'e}nez, Pablo and Serra, Xavier and Bogdanov, Dmitry},\n  booktitle={Conference of the International Society for Music Information Retrieval (ISMIR)},\n  year={2022}\n}",
    "inference": {
        "sample_rate": 16000,
        "algorithm": "TensorflowPredict2D",
        "embedding_model": {
            "algorithm": "TensorflowPredictEffnetDiscogs",
            "model_name": "discogs-effnet-bs64-1",
            "link": "https://essentia.upf.edu/models/music-style-classification/discogs-effnet/discogs-effnet-bs64-1.pb"
        }
    }
}