shadow-cann/hispark-modelzoo-yolov7

YOLOv7

YOLOv7在速度与精度方面均超越现有已知目标检测器：在5-160 FPS范围内表现最优，并在GPU V100上以30+ FPS实现56.8% AP的最高精度。其YOLOv7-E6模型在V100上达到56 FPS和55.9% AP，相比基于Transformer的SWIN-L Cascade-Mask R-CNN（A100 9.2 FPS，53.9% AP）速度提升509%且精度提高2%；相较基于卷积的ConvNeXt-XL Cascade-Mask R-CNN（A100 8.6 FPS，55.2% AP）速度提升551%且精度提高0.7%。此外，YOLOv7在速度与精度上均优于YOLOR、YOLOX、YOLOv5等主流检测器，且仅使用MS COCO数据集从头训练，未借助任何预训练权重。

Mirror Metadata

• Hugging Face repo: shadow-cann/hispark-modelzoo-yolov7
• Portal model id: hs9rrefl5c00
• Created at: 2025-11-14 10:24:41
• Updated at: 2025-11-29 14:49:58
• Category: 计算机视觉

Framework

• PyTorch

Supported OS

• OpenHarmony
• Linux

Computing Power

• Hi3403V100 SVP_NNN
• Hi3403V100 NNN

Tags

• 目标检测

Detail Parameters

• 输入: 640x640
• 参数量: 36.922M
• 计算量: 110.553 GFLOPs

Files In This Repo

• yolov7.pt (源模型 / 源模型下载; 源模型 / 源模型元数据)
• yolov7.onnx (源模型 / 源模型下载; 源模型 / 源模型元数据)
• SVP_NNN_PC_V1.0.6.0.tgz (附加资源 / 附加资源)

Upstream Links

• Portal card: https://gitbubble.github.io/hisilicon-developer-portal-mirror/model-detail.html?id=hs9rrefl5c00
• Upstream repository: https://gitee.com/HiSpark/modelzoo/tree/master/samples/samples_GPL/built-in/yolov7
• License reference: https://github.com/WongKinYiu/yolov7/blob/main/LICENSE.md

Notes

• This repository was mirrored from the HiSilicon Developer Portal model card and local downloads captured on 2026-03-27.
• File ownership follows the portal card mapping, not just filename similarity.
• Cover image: 1712178182881282_test067_output.jpg