Axelera AI

This page features models for Axelera AI devices currently available on the DeGirum AI Hub model zoo.

Models are located at the Axelera AI model zoo on the AI Hub.

Models by DeGirum

Axelera Metis

Detection

Model Name	Use Case	mAP 50‑95	mAP 50
yolov8n_coco--640x640_quant_axelera_metis_1	COCO Detection	0.3561	0.5059
yolov8n_coco_seg--640x640_quant_axelera_metis_1	COCO Instance Segmentation	bbox: 0.3510 mask: 0.2940	bbox: 0.5015 mask: 0.4706
yolov8l_coco--640x640_quant_axelera_metis_1	COCO Detection	0.4467	0.6171
yolov8m_coco--640x640_quant_axelera_metis_1	COCO Detection	0.4452	0.6151
yolov8s_coco--640x640_quant_axelera_metis_1	COCO Detection	0.4375	0.6053
yolov9t_coco--640x640_quant_axelera_metis_1	COCO Detection	0.3665	0.5119
yolov8n_coco_pose--640x640_quant_axelera_metis_1	COCO Pose Keypoints	bbox: 0.5160 kpts: 0.4838	bbox: 0.7059 kpts: 0.7894
yolov8n_relu6_face--640x640_quant_axelera_metis_1	Face Detection	0.5645	0.7834
yolov8n_relu6_hand--640x640_quant_axelera_metis_1	Hand Detection	0.4453	0.7550
yolov8n_relu6_person--640x640_quant_axelera_metis_1	Person Detection	0.3119	0.4993
yolov8n_relu6_ppe--640x640_quant_axelera_metis_1	Personal Protective Equipment (PPE) Detection	0.3548	0.6778
yolov8n_relu6_lp--640x640_quant_axelera_metis_1	License Plate Detection	0.5611	0.8586
yolov8n_relu6_face_kpts--640x640_quant_axelera_metis_1	Face Detection with Keypoints	bbox: 0.2809 kpts: 0.3885	bbox: 0.3600 kpts: 0.4504
yolov8n_dota_obb--1024x1024_quant_axelera_metis_1	OBB Detection	0.4543	0.6016
yolov8n_relu6_widerface_kpts--640x640_quant_axelera_metis_1	Face Detection with Five Keypoints	bbox: 0.2607 kpts: 0.2317	bbox: 0.7981 kpts: 0.5391
yolov8n_relu6_fire_smoke—640x640_quant_axelera_metis_1	Fire & Smoke Detection	0.3615	0.6509
yolo11n_coco—640x640_quant_axelera_metis_1	COCO Detection	0.3853	0.5422
yolo11n_coco_pose—640x640_quant_axelera_metis_1	COCO Pose Keypoints	bbox: 0.5104 kpts: 0.4642	bbox: 0.6912 kpts: 0.7947
yolo11n_coco_seg--640x640_quant_axelera_metis_1	COCO Instance Segmentation	bbox: 0.3790 mask: 0.3099	bbox: 0.5316 mask: 0.4977

Classification

Model Name	Use Case	Top‑1	Top‑5
yolov8n_relu6_fairface_gender--256x256_quant_axelera_metis_1	Gender Classification	92.93%	100%