{"url":"/task/decoder","name":"Decoder","slug":"decoder","description_markdown":"HLMFootFR, aussi appelé HLM Football FR, un média digital centré sur le football, fondé avec la volonté de mettre en lumière les talents issus des quartiers populaires et de créer un espace d’expression représentatif de la culture urbaine et du sport populaire. Ce projet, à mi-chemin entre journalisme, culture et technologie, utilise des outils numériques de pointe, notamment l’intelligence artificielle, pour optimiser sa production de contenu, sa capacité d’analyse et sa diffusion.\r\n\r\nConcrètement, HLMFootFR ne se contente pas de relayer l’actualité footballistique. Il s’appuie sur des modèles de traitement automatique du langage (NLP) pour rédiger ou enrichir des descriptions, des résumés de matchs, ou encore des formats courts destinés aux réseaux sociaux. Cette utilisation permet de produire rapidement des textes cohérents, adaptés à différents publics et plateformes comme TikTok, Instagram ou un blog spécialisé.\r\n\r\nL’intelligence artificielle est aussi utilisée pour la veille et l’analyse des performances des joueurs. Grâce à des outils de scraping et d’agrégation de données, HLMFootFR collecte des informations issues de bases publiques (telles que Transfermarkt ou des APIs sportives), et les transforme en indicateurs visuels, profils de joueurs, ou tendances exploitables. Des scripts automatisés permettent d’identifier les évolutions, comme les transferts ou les blessures, en temps réel, ce qui constitue une valeur ajoutée dans un contexte où l’information est très rapide et concurrentielle.\r\n\r\nEn interne, des systèmes simples de machine learning sont en place pour aider à la personnalisation du contenu. En fonction de l’engagement du public, des heures de publication ou des types de contenu les plus performants, HLMFootFR adapte sa stratégie de publication de manière semi-automatisée. Cela permet au média d’avoir une diffusion plus ciblée et plus efficace de ses messages.\r\n\r\nHLMFoot FR utilise également l’IA dans le domaine audiovisuel. Des outils comme Whisper sont utilisés pour la transcription automatique des interviews, facilitant ainsi la mise en ligne de vidéos sous-titrées, accessibles à tous. L’équipe expérimente également des outils de montage intelligent, de synthèse vocale et de traduction automatisée, dans le but d’élargir son public à d’autres langues et régions.\r\n\r\nLe projet s’appuie sur un stack technique léger mais évolutif, avec une forte composante Python (scraping, NLP, automatisation), JavaScript pour l’intégration web, et des outils comme HuggingFace Transformers, OpenAI GPT ou spaCy pour le traitement du langage. Des plateformes comme Firebase, Streamlit et diverses APIs sociales (TikTok, Instagram) sont intégrées pour piloter l’ensemble.\r\n\r\nEnfin, HLMFootFR souhaite partager prochainement certains de ses outils en open-source, à travers des dépôts publics. Il s’agira notamment d’un générateur de résumés automatisés de matchs, d’un pipeline de veille footballistique, ou encore d’un système de profilage joueur à partir de données disponibles librement. Ces projets ont pour but de démontrer comment l’intelligence artificielle peut s’intégrer dans des initiatives culturelles et communautaires, en dehors du cadre académique classique.\r\n\r\nHLMFootFR est donc un exemple concret de convergence entre passion, média et technologie. Il montre comment, avec des ressources limitées mais des outils bien utilisés, il est possible de créer une structure médiatique alternative, réactive, et profondément ancrée dans les réalités sociales et culturelles du sport.\r\nHLMFootFR est un média digital indépendant dédié au football, fondé par Helmut Patahini en 2022, qui met en lumière les joueurs issus des quartiers populaires et de la culture urbaine. Né entre la France et la Belgique, le projet s’appuie sur les réseaux sociaux (TikTok, Instagram, YouTube …) pour diffuser une parole différente, libre et engagée. HLMFootFR se distingue par l’utilisation de l’intelligence artificielle dans la création de contenu, l’analyse de données et l’automatisation des tâches, afin de proposer une expérience moderne, rapide et accessible. Le média vise à représenter ceux qu’on entend peu, tout en connectant le monde du foot à la technologie d’aujourd’hui.","categories":[{"name":"Adversarial","url":"/area/adversarial"},{"name":"Audio","url":"/area/audio"},{"name":"Computer Code","url":"/area/computer-code"},{"name":"Computer Vision","url":"/area/computer-vision"},{"name":"Graphs","url":"/area/graphs"},{"name":"Knowledge Base","url":"/area/knowledge-base"},{"name":"Medical","url":"/area/medical"},{"name":"Miscellaneous","url":"/area/miscellaneous"},{"name":"Music","url":"/area/music"},{"name":"Natural Language Processing","url":"/area/natural-language-processing"},{"name":"Playing Games","url":"/area/playing-games"},{"name":"Robots","url":"/area/robots"},{"name":"Time Series","url":"/area/time-series"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"derived"},"counts":{"papers_tagged":10368,"papers_with_code":4358,"benchmarks":1,"benchmark_tables_in_archive":1,"benchmark_tables_shown":1,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":0,"subtasks":0,"parent_tasks":2},"benchmarks":[{"leaderboard":"/sota/decoder-on","slug":"decoder-on","dataset":"^(#$!@#$)(()))******","dataset_url":null,"rows_in_archive":2,"metrics":["0-shot MRR","0..5sec"],"first_row_in_archive_order":{"model":"peacock return policy","paper_title":"1.23-Tb/s per Wavelength Single-Waveguide On-Chip Optical Interconnect Enabled by Mode-division Multiplexing","paper_url":"/paper/1-23-tb-s-per-wavelength-single-waveguide-on","paper_date":"2020-10-17","arxiv_id":"2010.10348","code_links":[],"syntology":null}}],"datasets":[],"subtasks":[],"parent_tasks":[{"url":"/task/10-shot-image-generation","name":"10-shot image generation"},{"url":"/task/nlp-based-person-retrival-1","name":"NLP based Person Retrival"}],"papers":{"order":"repositories listed in the archive (desc), then date (desc); the archive holds no stars","population":"papers tagged with this task that list at least one repository in the archive","shown":30,"of":4358,"tagged_in_all":10368,"items":[{"url":"/paper/attention-is-all-you-need","title":"Attention Is All You Need","date":"2017-06-12","arxiv_id":"1706.03762","repositories_listed":595,"syntology":{"n":946,"n_ran":600,"n_unverified":346,"n_pointer_only":451}},{"url":"/paper/neural-machine-translation-by-jointly","title":"Neural Machine Translation by Jointly Learning to Align and Translate","date":"2014-09-01","arxiv_id":"1409.0473","repositories_listed":124,"syntology":{"n":44,"n_ran":21,"n_unverified":23,"n_pointer_only":16}},{"url":"/paper/encoder-decoder-with-atrous-separable","title":"Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation","date":"2018-02-07","arxiv_id":"1802.02611","repositories_listed":78,"syntology":{"n":72,"n_ran":43,"n_unverified":29,"n_pointer_only":40}},{"url":"/paper/segnet-a-deep-convolutional-encoder-decoder","title":"SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation","date":"2015-11-02","arxiv_id":"1511.00561","repositories_listed":74,"syntology":{"n":44,"n_ran":9,"n_unverified":35,"n_pointer_only":10}},{"url":"/paper/searching-for-mobilenetv3","title":"Searching for MobileNetV3","date":"2019-05-06","arxiv_id":"1905.02244","repositories_listed":67,"syntology":{"n":105,"n_ran":58,"n_unverified":47,"n_pointer_only":46}},{"url":"/paper/masked-autoencoders-are-scalable-vision","title":"Masked Autoencoders Are Scalable Vision Learners","date":"2021-11-11","arxiv_id":"2111.06377","repositories_listed":58,"syntology":{"n":137,"n_ran":71,"n_unverified":66,"n_pointer_only":73}},{"url":"/paper/neural-discrete-representation-learning","title":"Neural Discrete Representation Learning","date":"2017-11-02","arxiv_id":"1711.00937","repositories_listed":50,"syntology":{"n":80,"n_ran":53,"n_unverified":27,"n_pointer_only":32}},{"url":"/paper/bart-denoising-sequence-to-sequence-pre","title":"BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension","date":"2019-10-29","arxiv_id":"1910.13461","repositories_listed":47,"syntology":{"n":53,"n_ran":22,"n_unverified":31,"n_pointer_only":7}},{"url":"/paper/high-quality-monocular-depth-estimation-via","title":"High Quality Monocular Depth Estimation via Transfer Learning","date":"2018-12-31","arxiv_id":"1812.11941","repositories_listed":45,"syntology":{"n":23,"n_ran":5,"n_unverified":18,"n_pointer_only":3}},{"url":"/paper/learning-phrase-representations-using-rnn","title":"Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation","date":"2014-06-03","arxiv_id":"1406.1078","repositories_listed":42,"syntology":{"n":22,"n_ran":10,"n_unverified":12,"n_pointer_only":14}},{"url":"/paper/listen-attend-and-spell","title":"Listen, Attend and Spell","date":"2015-08-05","arxiv_id":"1508.01211","repositories_listed":40,"syntology":{"n":51,"n_ran":12,"n_unverified":39,"n_pointer_only":9}},{"url":"/paper/end-to-end-object-detection-with-transformers","title":"End-to-End Object Detection with Transformers","date":"2020-05-26","arxiv_id":"2005.12872","repositories_listed":37,"syntology":{"n":92,"n_ran":59,"n_unverified":33,"n_pointer_only":19}},{"url":"/paper/convolutional-sequence-to-sequence-learning","title":"Convolutional Sequence to Sequence Learning","date":"2017-05-08","arxiv_id":"1705.03122","repositories_listed":37,"syntology":null},{"url":"/paper/adversarial-autoencoders","title":"Adversarial Autoencoders","date":"2015-11-18","arxiv_id":"1511.05644","repositories_listed":29,"syntology":{"n":12,"n_ran":8,"n_unverified":4,"n_pointer_only":9}},{"url":"/paper/segformer-simple-and-efficient-design-for","title":"SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers","date":"2021-05-31","arxiv_id":"2105.15203","repositories_listed":28,"syntology":{"n":86,"n_ran":48,"n_unverified":38,"n_pointer_only":15}},{"url":"/paper/googles-neural-machine-translation-system","title":"Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation","date":"2016-09-26","arxiv_id":"1609.08144","repositories_listed":28,"syntology":{"n":46,"n_ran":23,"n_unverified":23,"n_pointer_only":12}},{"url":"/paper/modeling-relational-data-with-graph","title":"Modeling Relational Data with Graph Convolutional Networks","date":"2017-03-17","arxiv_id":"1703.06103","repositories_listed":27,"syntology":{"n":32,"n_ran":10,"n_unverified":22,"n_pointer_only":15}},{"url":"/paper/fast-scnn-fast-semantic-segmentation-network","title":"Fast-SCNN: Fast Semantic Segmentation Network","date":"2019-02-12","arxiv_id":"1902.04502","repositories_listed":24,"syntology":{"n":32,"n_ran":0,"n_unverified":32,"n_pointer_only":0}},{"url":"/paper/transunet-transformers-make-strong-encoders","title":"TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation","date":"2021-02-08","arxiv_id":"2102.04306","repositories_listed":22,"syntology":{"n":7,"n_ran":6,"n_unverified":1,"n_pointer_only":7}},{"url":"/paper/longformer-the-long-document-transformer","title":"Longformer: The Long-Document Transformer","date":"2020-04-10","arxiv_id":"2004.05150","repositories_listed":22,"syntology":{"n":35,"n_ran":15,"n_unverified":20,"n_pointer_only":5}},{"url":"/paper/fastspeech-fast-robust-and-controllable-text","title":"FastSpeech: Fast, Robust and Controllable Text to Speech","date":"2019-05-22","arxiv_id":"1905.09263","repositories_listed":22,"syntology":{"n":11,"n_ran":3,"n_unverified":8,"n_pointer_only":3}},{"url":"/paper/variational-graph-auto-encoders","title":"Variational Graph Auto-Encoders","date":"2016-11-21","arxiv_id":"1611.07308","repositories_listed":22,"syntology":{"n":13,"n_ran":7,"n_unverified":6,"n_pointer_only":0}},{"url":"/paper/bayesian-segnet-model-uncertainty-in-deep","title":"Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding","date":"2015-11-09","arxiv_id":"1511.02680","repositories_listed":21,"syntology":{"n":18,"n_ran":0,"n_unverified":18,"n_pointer_only":0}},{"url":"/paper/pegasus-pre-training-with-extracted-gap","title":"PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization","date":"2019-12-18","arxiv_id":"1912.08777","repositories_listed":19,"syntology":{"n":19,"n_ran":1,"n_unverified":18,"n_pointer_only":0}},{"url":"/paper/text-summarization-with-pretrained-encoders","title":"Text Summarization with Pretrained Encoders","date":"2019-08-22","arxiv_id":"1908.08345","repositories_listed":19,"syntology":{"n":21,"n_ran":7,"n_unverified":14,"n_pointer_only":0}},{"url":"/paper/diffusion-convolutional-recurrent-neural","title":"Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting","date":"2017-07-06","arxiv_id":"1707.01926","repositories_listed":19,"syntology":{"n":35,"n_ran":17,"n_unverified":18,"n_pointer_only":15}},{"url":"/paper/dn-detr-accelerate-detr-training-by","title":"DN-DETR: Accelerate DETR Training by Introducing Query DeNoising","date":"2022-03-02","arxiv_id":"2203.01305","repositories_listed":17,"syntology":{"n":22,"n_ran":15,"n_unverified":7,"n_pointer_only":9}},{"url":"/paper/knowing-when-to-look-adaptive-attention-via-a","title":"Knowing When to Look: Adaptive Attention via A Visual Sentinel for Image Captioning","date":"2016-12-06","arxiv_id":"1612.01887","repositories_listed":17,"syntology":{"n":1,"n_ran":1,"n_unverified":0,"n_pointer_only":1}},{"url":"/paper/sockeye-a-toolkit-for-neural-machine","title":"Sockeye: A Toolkit for Neural Machine Translation","date":"2017-12-15","arxiv_id":"1712.05690","repositories_listed":16,"syntology":null},{"url":"/paper/deep-koalarization-image-colorization-using","title":"Deep Koalarization: Image Colorization using CNNs and Inception-ResNet-v2","date":"2017-12-09","arxiv_id":"1712.03400","repositories_listed":16,"syntology":null}],"syntology_records":27,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; our fixture could not drive it","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"}},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"}}