NEWVectors or files. Pick a path.Start →
    HFObject Detectionapache-2.0

    yolos-tiny

    by hustvl

    You Only Look at One Sequence, ViT-based real-time object detection

    147Kdl/month
    282likes
    6Mparams
    Identifiers
    Model ID
    hustvl/yolos-tiny
    Feature URI
    mixpeek://image_extractor@v1/hustvl_yolos_tiny_v1

    Overview

    YOLOS adapts the Vision Transformer (ViT) architecture for object detection by simply appending detection tokens to the input sequence. It demonstrates that a pure transformer can perform object detection without any convolutional components.

    On Mixpeek, YOLOS Tiny provides a lightweight, fast alternative to DETR for object detection tasks where speed is prioritized over maximum accuracy.

    Architecture

    Vision Transformer (ViT-Tiny) with 12 layers. Appends 100 learnable detection tokens to the image patch sequence. Uses bipartite matching loss like DETR.

    Mixpeek SDK Integration

    import { Mixpeek } from "mixpeek";
    
    const mx = new Mixpeek({ apiKey: "API_KEY" });
    
    // Managed: create a collection over a bucket; Mixpeek runs this model's extractor
    const collection = await mx.collections.create({
      namespace_id: "my-namespace",
      collection_name: "my-collection",
      source: { type: "bucket", bucket_ids: ["bkt_your_bucket"] },
      feature_extractor: {
        feature_extractor_name: "object_detection",
        version: "v1",
        parameters: { model_id: "hustvl/yolos-tiny" },
      },
    });

    Capabilities

    • Lightweight ViT-based object detection
    • Fast inference suitable for real-time processing
    • COCO object categories
    • Pure transformer architecture (no CNN backbone)

    Use Cases on Mixpeek

    Real-time video analysis where low latency is critical
    Edge deployment scenarios with limited compute
    High-throughput batch processing of large video archives

    Benchmarks

    DatasetMetricScoreSource
    COCO val2017AP (box)30.4Fang et al., 2021 — Table 1
    COCO val2017AP5048.6Fang et al., 2021 — Table 1

    Performance

    Input Size512×864 px
    GPU Latency~6ms / image (A100)
    CPU Latency~55ms / image
    GPU Throughput~165 images/sec (A100)
    GPU Memory~0.4 GB

    6.5M params — optimized for edge and high-throughput scenarios

    Specification

    FrameworkHF
    Organizationhustvl
    FeatureObject Detection
    Outputbbox + label
    Modalitiesvideo, image
    RetrieverObject Filter
    Parameters6M
    Licenseapache-2.0
    Downloads/mo147K
    Likes282

    Research Paper

    You Only Look at One Sequence

    arxiv.org

    Build a pipeline with yolos-tiny

    Add this model to a processing pipeline alongside other extractors. Combine with retrieval stages for end-to-end search.

    Run on your data, free