MediaPipe Solutions proporciona un conjunto de bibliotecas y herramientas para que apliques rápidamente técnicas de inteligencia artificial (IA) y aprendizaje automático (AA) en tus aplicaciones. Puedes conectar estas soluciones a tus aplicaciones de inmediato, personalizarlas según tus necesidades y usarlas en múltiples plataformas de desarrollo. Las soluciones de MediaPipe forman parte del proyecto de código abierto de MediaPipe, por lo que puedes personalizar aún más el código de las soluciones para satisfacer las necesidades de tu aplicación. El paquete de soluciones de MediaPipe incluye lo siguiente:
Estas bibliotecas y recursos proporcionan la funcionalidad principal para cada solución de MediaPipe:
MediaPipe Tasks: APIs y bibliotecas multiplataforma para implementar soluciones. Más información
Modelos de MediaPipe: Modelos previamente entrenados y listos para ejecutar que se pueden usar con cada solución.
Estas herramientas te permiten personalizar y evaluar soluciones:
MediaPipe Model Maker: Personaliza modelos para soluciones con tus datos. Más información
MediaPipe Studio: Visualiza, evalúa y compara soluciones en tu navegador. Más información
Soluciones disponibles
Las MediaPipe Solutions están disponibles en varias plataformas. Cada solución incluye uno o más modelos, y también puedes personalizar modelos para algunas soluciones. En la siguiente lista, se muestran las soluciones disponibles para cada plataforma compatible y si puedes usar Model Maker para personalizar el modelo:
Para comenzar a usar las MediaPipe Solutions, selecciona cualquiera de las tareas que se indican en el árbol de navegación de la izquierda, incluidas las tareas de visión, texto y audio.
Si necesitas ayuda para configurar un entorno de desarrollo para usar con las tareas de MediaPipe, consulta las guías de configuración para Android, apps web y Python.
Soluciones heredadas
Desde el 1 de marzo de 2023, ya no brindamos asistencia para las soluciones heredadas de MediaPipe que se indican a continuación. Todas las demás soluciones heredadas de MediaPipe se actualizarán a una nueva solución de MediaPipe. Consulta la siguiente lista para obtener más detalles. El repositorio de código y los archivos binarios prediseñados para todas las soluciones heredadas de MediaPipe se seguirán proporcionando tal como están.
[[["Fácil de comprender","easyToUnderstand","thumb-up"],["Resolvió mi problema","solvedMyProblem","thumb-up"],["Otro","otherUp","thumb-up"]],[["Falta la información que necesito","missingTheInformationINeed","thumb-down"],["Muy complicado o demasiados pasos","tooComplicatedTooManySteps","thumb-down"],["Desactualizado","outOfDate","thumb-down"],["Problema de traducción","translationIssue","thumb-down"],["Problema con las muestras o los códigos","samplesCodeIssue","thumb-down"],["Otro","otherDown","thumb-down"]],["Última actualización: 2025-09-09 (UTC)"],[],[],null,["MediaPipe Solutions provides a suite of libraries and tools for you to quickly\napply artificial intelligence (AI) and machine learning (ML) techniques in your\napplications. You can plug these solutions into your applications immediately,\ncustomize them to your needs, and use them across multiple development\nplatforms. MediaPipe Solutions is part of the MediaPipe [open source\nproject](https://github.com/google/mediapipe), so you can further customize the\nsolutions code to meet your application needs. The MediaPipe Solutions suite\nincludes the following:\n\nThese libraries and resources provide the core functionality for each MediaPipe\nSolution:\n\n- **MediaPipe Tasks** : Cross-platform APIs and libraries for deploying solutions. [Learn more](/edge/mediapipe/solutions/tasks)\n- **MediaPipe Models**: Pre-trained, ready-to-run models for use with each solution.\n\nThese tools let you customize and evaluate solutions:\n\n- **MediaPipe Model Maker** : Customize models for solutions with your data. [Learn\n more](/edge/mediapipe/solutions/model_maker)\n- **MediaPipe Studio** : Visualize, evaluate, and benchmark solutions in your browser. [Learn\n more](/edge/mediapipe/solutions/studio)\n\nAvailable solutions\n\nMediaPipe Solutions are available across multiple platforms. Each solution\nincludes one or more models, and you can customize models for some solutions as\nwell. The following list shows what solutions are available for each supported\nplatform and if you can use Model Maker to customize the model:\n\n| Solution | Android | Web | Python | iOS | Customize model |\n|------------------------------------------------------------------------------------|---------|-----|--------|-----|-----------------|\n| [LLM Inference API](/edge/mediapipe/solutions/genai/llm_inference) | | | | | |\n| [Object detection](/edge/mediapipe/solutions/vision/object_detector) | | | | | |\n| [Image classification](/edge/mediapipe/solutions/vision/image_classifier) | | | | | |\n| [Image segmentation](/edge/mediapipe/solutions/vision/image_segmenter) | | | | | |\n| [Interactive segmentation](/edge/mediapipe/solutions/vision/interactive_segmenter) | | | | | |\n| [Hand landmark detection](/edge/mediapipe/solutions/vision/hand_landmarker) | | | | | |\n| [Gesture recognition](/edge/mediapipe/solutions/vision/gesture_recognizer) | | | | | |\n| [Image embedding](/edge/mediapipe/solutions/vision/image_embedder) | | | | | |\n| [Face detection](/edge/mediapipe/solutions/vision/face_detector) | | | | | |\n| [Face landmark detection](/edge/mediapipe/solutions/vision/face_landmarker) | | | | | |\n| [Pose landmark detection](/edge/mediapipe/solutions/vision/pose_landmarker) | | | | | |\n| [Image generation](/edge/mediapipe/solutions/vision/image_generator) | | | | | |\n| [Text classification](/edge/mediapipe/solutions/text/text_classifier) | | | | | |\n| [Text embedding](/edge/mediapipe/solutions/text/text_embedder) | | | | | |\n| [Language detector](/edge/mediapipe/solutions/text/language_detector) | | | | | |\n| [Audio classification](/edge/mediapipe/solutions/audio/audio_classifier) | | | | | |\n\nGet started\n\nYou can get started with MediaPipe Solutions by selecting any of the tasks\nlisted in the left navigation tree, including\n[vision](/edge/mediapipe/solutions/vision/object_detector),\n[text](/edge/mediapipe/solutions/text/text_classifier), and\n[audio](/edge/mediapipe/solutions/audio/audio_classifier) tasks.\nIf you need help setting up a development environment for use with MediaPipe\nTasks, check out the setup guides for\n[Android](/edge/mediapipe/solutions/setup_android),\n[web apps](/edge/mediapipe/solutions/setup_web), and\n[Python](/edge/mediapipe/solutions/setup_python).\n\nLegacy solutions\n\nWe have ended support for the MediaPipe Legacy Solutions listed below as of\nMarch 1, 2023. All other MediaPipe Legacy Solutions will be upgraded to a new\nMediaPipe Solution. See the list below for details. The\n[code repository](https://github.com/google/mediapipe/tree/master/mediapipe) and\nprebuilt binaries for all MediaPipe Legacy Solutions will continue to be\nprovided on an as-is basis.\n\n| Legacy Solution | Status | New MediaPipe Solution |\n|-----------------------------------------------------------------------------------------------------------------------------|--------------------------------------|--------------------------------------------------------------|\n| Face Detection ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/face_detection.md)) | [Upgraded](./vision/face_detector) | [Face detection](./vision/face_detector) |\n| Face Mesh ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/face_mesh.md)) | [Upgraded](./vision/face_landmarker) | [Face landmark detection](./vision/face_landmarker) |\n| Iris ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/iris.md)) | [Upgraded](./vision/face_landmarker) | [Face landmark detection](./vision/face_landmarker) |\n| Hands ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/hands.md)) | [Upgraded](./vision/hand_landmarker) | [Hand landmark detection](./vision/hand_landmarker) |\n| Pose ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/pose.md)) | [Upgraded](./vision/pose_landmarker) | [Pose landmark detection](./vision/pose_landmarker) |\n| Holistic ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/holistic.md)) | Upgrade | [Holistic landmarks detection](./vision/holistic_landmarker) |\n| Selfie segmentation ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/selfie_segmentation.md)) | [Upgraded](./vision/image_segmenter) | [Image segmentation](./vision/image_segmenter) |\n| Hair segmentation ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/hair_segmentation.md)) | [Upgraded](./vision/image_segmenter) | [Image segmentation](./vision/image_segmenter) |\n| Object detection ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/object_detection.md)) | [Upgraded](./vision/object_detector) | [Object detection](./vision/object_detector) |\n| Box tracking ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/box_tracking.md)) | Support ended | |\n| Instant motion tracking ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/instant_motion_tracking.md)) | Support ended | |\n| Objectron ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/objectron.md)) | Support ended | |\n| KNIFT ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/knift.md)) | Support ended | |\n| AutoFlip ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/autoflip.md)) | Support ended | |\n| MediaSequence ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/media_sequence.md)) | Support ended | |\n| YouTube 8M ([info](https://github.com/google/mediapipe/blob/master/docs/solutions/youtube_8m.md)) | Support ended | |"]]