awesome-repositories.com
Blog
MCP
awesome-repositories.com

Découvrez les meilleurs dépôts open-source grâce à notre recherche par IA.

ExplorerRecherches sélectionnéesAlternatives open sourceLogiciels auto-hébergésBlogPlan du site
ProjetServeur MCPÀ proposNotre méthodologiePresse
Mentions légalesConfidentialitéConditions d'utilisation
© 2026 Bringes Technology SRL·VAT RO45896025·hello@awesome-repositories.com
·

9 dépôts

Awesome GitHub RepositoriesSpatial Downsampling

Techniques for reducing the spatial resolution of feature maps in neural networks via pooling or strided convolutions.

Distinguishing note: Candidates refer to time-series metric downsampling or cache limits, not CNN spatial resolution reduction.

Explore 9 awesome GitHub repositories matching artificial intelligence & ml · Spatial Downsampling. Refine with filters or upvote what's useful.

Awesome Spatial Downsampling GitHub Repositories

Trouvez les meilleurs dépôts grâce à l'IA.Nous recherchons les dépôts les plus pertinents grâce à l'IA.
  • exacity/deeplearningbook-chineseAvatar de exacity

    exacity/deeplearningbook-chinese

    37,285Voir sur GitHub↗

    This project is a comprehensive Chinese translation of a technical deep learning textbook, providing an educational resource on the theory and implementation of neural networks. It functions as a collaborative technical translation project designed to make complex academic AI literature accessible to non-English speakers. The project utilizes a community-driven translation model that integrates external suggestions and pull requests to refine linguistic accuracy and reduce bias. It employs standardized terminology mapping to ensure a uniform vocabulary throughout the translated content. To i

    Discusses reducing the spatial size of feature maps by skipping positions during convolution.

    TeX
    Voir sur GitHub↗37,285
  • lucidrains/vit-pytorchAvatar de lucidrains

    lucidrains/vit-pytorch

    25,363Voir sur GitHub↗

    This library provides a comprehensive collection of modular building blocks and research-backed architectures for implementing vision transformers within the PyTorch framework. It serves as a centralized repository for constructing, training, and analyzing attention-based models, offering a wide array of specialized variants designed for image classification and visual representation learning. The project distinguishes itself through a focus on architectural efficiency and flexibility, supporting diverse input formats including non-square images and volumetric data like video. It incorporates

    Reduces sequence length through unfolding or pooling operations to capture hierarchical features within transformer layers.

    Python
    Voir sur GitHub↗25,363
  • improbable-eng/thanosAvatar de improbable-eng

    improbable-eng/thanos

    14,105Voir sur GitHub↗

    Thanos is a CNCF cloud native monitoring tool that provides a highly available and scalable extension to the Prometheus ecosystem. It functions as a global query engine, a long-term storage system, and a metric downsampler. The project enables a unified interface to aggregate and query metrics across multiple distributed clusters from a single view. It maintains historical data beyond local retention limits by persisting time-series metrics in object storage and eliminates data gaps by merging metrics from redundant server pairs. The system includes capabilities for reducing the resolution o

    Creates lower-resolution versions of historical metrics to accelerate queries over massive datasets.

    Go
    Voir sur GitHub↗14,105
  • jackzhenguo/python-small-examplesAvatar de jackzhenguo

    jackzhenguo/python-small-examples

    8,132Voir sur GitHub↗

    This project is a comprehensive library of practical Python code examples and patterns. It provides a collection of scripts and snippets designed to demonstrate a wide range of programming tasks, from basic syntax to advanced implementation patterns. The repository focuses on several core domains, including the implementation of concurrency and multithreading examples, data analysis snippets for cleaning and manipulating tabular data, and various data visualization examples. It also covers automation scripts for file system management and a variety of general programming patterns. Additional

    Provides a tool to aggregate high-frequency time series data into lower-frequency intervals.

    Pythondata-sciencemachine-learningpython
    Voir sur GitHub↗8,132
  • open-mmlab/mmagicAvatar de open-mmlab

    open-mmlab/mmagic

    7,434Voir sur GitHub↗

    mmagic is a multimodal training pipeline and framework for generative AI, focusing on visual synthesis and restoration. It provides the infrastructure to build and train models for tasks such as text-to-image and text-to-video generation, 3D-aware content synthesis, and high-fidelity image translation using diffusion models and generative adversarial networks. The project distinguishes itself through specialized capabilities for generative model personalization, including techniques for fine-tuning subjects and styles. It also supports advanced visual manipulations such as latent space interp

    Implements pixel unshuffle downsampling to reduce spatial resolution by rearranging pixels into the channel dimension.

    Jupyter Notebookaigccomputer-visiondeep-learning
    Voir sur GitHub↗7,434
  • instillai/machine-learning-courseAvatar de instillai

    instillai/machine-learning-course

    7,043Voir sur GitHub↗

    Il s'agit d'un programme éducatif complet conçu pour enseigner les fondamentaux de l'apprentissage automatique en utilisant le langage de programmation Python. Il fournit un cours structuré couvrant l'implémentation et la théorie de l'apprentissage supervisé, de l'apprentissage non supervisé et de l'apprentissage profond. Le programme est dispensé via des notebooks interactifs qui combinent du code exécutable avec des tutoriels techniques. Il inclut des guides dédiés pour construire des architectures de réseaux de neurones, implémenter des modèles de classification et de régression, et utiliser des techniques de clustering pour la découverte de modèles dans des données non étiquetées. Le matériel couvre l'intégralité du flux de travail d'apprentissage automatique, y compris le prétraitement des données et l'encodage catégoriel, l'entraînement des modèles et le réglage des hyperparamètres, ainsi que l'évaluation des performances. Il propose également des outils pour visualiser le comportement des modèles, tels que le tracé des frontières de décision et les diagrammes d'arbres de décision.

    Provides utilities for downsampling feature maps using max pooling to reduce computational overhead and noise.

    Python
    Voir sur GitHub↗7,043
  • facebookresearch/convnextAvatar de facebookresearch

    facebookresearch/ConvNeXt

    6,388Voir sur GitHub↗

    Code release for ConvNeXt model

    Implements strided 2x2 convolutions for spatial downsampling in the ConvNeXt architecture.

    Python
    Voir sur GitHub↗6,388
  • greptimeteam/greptimedbAvatar de GreptimeTeam

    GreptimeTeam/greptimedb

    5,968Voir sur GitHub↗

    GreptimeDB is a distributed, open-source time-series database built for unified observability. It stores and queries metrics, logs, and traces together in a single columnar engine, supporting both SQL and PromQL for analysis. The database is designed as a Kubernetes-native operator with a decoupled compute and storage architecture, enabling horizontal scaling and multi-region deployment. What distinguishes GreptimeDB is its role as a multi-protocol ingestion gateway, accepting data through OpenTelemetry, Prometheus Remote Write, InfluxDB, Loki, Elasticsearch, Kafka, and MQTT protocols without

    Reduces data volume by applying operations like average pooling to stored time-series data for efficient analysis.

    Rustanalyticscloud-nativedatabase
    Voir sur GitHub↗5,968
  • opentsdb/opentsdbAvatar de OpenTSDB

    OpenTSDB/opentsdb

    5,068Voir sur GitHub↗

    OpenTSDB est une base de données de séries temporelles distribuée et un moteur de métriques conçu pour stocker et gérer des volumes massifs de métriques système à haute cardinalité. Il fonctionne comme un magasin de données et une plateforme d'analyse qui permet l'ingestion de métriques à grande échelle et la surveillance de la performance de l'infrastructure à travers un cluster distribué. Le système se distingue par une abstraction de stockage distribué qui supporte de multiples backends tels que HBase, Cassandra et Google Bigtable. Il utilise un arbre de métriques hiérarchique pour organiser les séries temporelles et emploie l'indexation par identifiant numérique pour réduire l'empreinte de stockage et accélérer les recherches pour les métriques taguées. Le projet couvre de larges domaines de capacités incluant l'analyse de données de séries temporelles avec des calculs de centiles distribués et le downsampling, ainsi qu'une gestion complète des métadonnées. Il fournit une intégration API pour l'ingestion et l'interrogation de données, le cache off-heap pour l'optimisation des performances, et des outils pour l'audit d'intégrité des données et l'analyse d'anomalies. Le système est géré via une interface en ligne de commande pour l'administration de la base de données et la synchronisation de l'arbre de métriques.

    Reduces the resolution of high-frequency time series data to optimize historical query performance.

    Java
    Voir sur GitHub↗5,068
  1. Home
  2. Artificial Intelligence & ML
  3. Spatial Downsampling

Explorer les sous-tags

  • Dimension-Matching ShortcutsSkip connections that use 1x1 convolutions to align tensor shapes for additive operations. **Distinct from Spatial Downsampling:** Focuses on the shape alignment mechanism for shortcuts, unlike general spatial downsampling.
  • Pixel UnshufflingA specific spatial downsampling method that rearranges pixels into the channel dimension to reduce resolution. **Distinct from Spatial Downsampling:** Distinct from general spatial downsampling as it specifically implements the unshuffle operation to preserve information in the channel dimension.
  • Strided Convolution DownsamplersDownsampling layers that use strided convolutions instead of pooling to reduce spatial dimensions. **Distinct from Spatial Downsampling:** Distinct from Spatial Downsampling: specifies strided convolution as the mechanism, not pooling or other methods.
  • Time-Series DownsamplersMechanisms for reducing the resolution of historical time-series data to improve query performance. **Distinct from Spatial Downsampling:** Focuses on temporal resolution reduction for metrics, distinct from spatial resolution reduction in neural networks.