CV / YOLACT

YOLACT-RegNetX-800MF

YOLACT (UC Davis) predicts instance masks with 49,104 anchors, Fast NMS and prototype-mask assembly (proto @ coeffs) on Hailo-8. Post-processing runs on the host CPU since the HEF exposes 16 raw heads.

1 downloads

Grootte

35 MB

Geheugen

4GB+

Precisie

Hailo HEF / HailoRT 4.23.x

Kies het apparaat dat je gebruikt. De installatiehandleiding en documentatie worden dienovereenkomstig bijgewerkt.

Platform

Inferentie-engine

Aan de slag

Implementeren

sudo docker run --rm \
  --name cm5-hailo8-yolact-800mf \
  --privileged \
  --net=host \
  -e PYTHONUNBUFFERED=1 \
  --device /dev/hailo0:/dev/hailo0 \
  -v /usr/lib/libhailort.so.4.23.0:/usr/lib/libhailort.so.4.23.0:ro \
  -v /usr/lib/libhailort.so:/usr/lib/libhailort.so:ro \
  ghcr.io/seeed-projects/recomputer-hailo8-cv/yolact_regnetx_800mf:latest \
  python web_detection.py --model_path model/yolact_regnetx_800mf.hef --video_path video/test.mp4

REST API

Gebruik de REST API om inferentie uit te voeren. Kopieer de onderstaande commando's.

Curl

curl -X POST "http://<Board_IP>:8000/api/models/yolact_regnetx_800mf/predict" \
  -F "file=@test.jpg"

Python

import requests

response = requests.post(
    "http://<Board_IP>:8000/api/models/yolact_regnetx_800mf/predict",
    files={"file": open("test.jpg", "rb")},
    timeout=30,
)
print(response.json())

Modeldetails

YOLACT-RegNetX-800MF on reComputer R Series (CM5 + Hailo-8)

YOLACT (UC Davis) performs real-time instance segmentation — per-object masks on top of COCO 80-class boxes — on Hailo-8 through HailoRT. The HEF exposes 16 raw heads and all decode runs on the host CPU.

Model Info

Property	Value
Architecture	YOLACT (You Only Look At CoefficienTs), RegNetX-800MF backbone
Task	Instance segmentation
Input	512x512x3 RGB (normalize_in_net ImageNet RGB mean/std)
Output	16 heads: proto (128x128x32) + 5 scales x {bbox, mask coeffs, conf}
Priors	49,104 anchors (9/cell, feature maps 64/32/16/8/4)
Classes	80 (COCO)
Parameters	28.3M
Operations	116.75G
mAP (COCO, full precision)	25.6% hardware reference
HEF	Hailo Model Zoo v2.19.0, Hailo-8

The accuracy value above is Hailo's Model Zoo reference. It is not a benchmark measured on CM5.

Run With Demo Video

bash

sudo docker run --rm \
  --name cm5-hailo8-yolact-800mf \
  --privileged \
  --net=host \
  -e PYTHONUNBUFFERED=1 \
  --device /dev/hailo0:/dev/hailo0 \
  -v /usr/lib/libhailort.so.4.23.0:/usr/lib/libhailort.so.4.23.0:ro \
  -v /usr/lib/libhailort.so:/usr/lib/libhailort.so:ro \
  ghcr.io/seeed-projects/recomputer-hailo8-cv/yolact_regnetx_800mf:latest \
  python web_detection.py --model_path model/yolact_regnetx_800mf.hef --video_path video/test.mp4

Open http://<Board_IP>:8000 to view the web preview (colored instance masks overlaid on the MJPEG stream).

REST API

bash

curl -X POST "http://<Board_IP>:8000/api/models/yolact_regnetx_800mf/predict" \
  -F "file=@test.jpg"

Endpoint	Method	Purpose
`/`	GET	Web preview UI
`/api/models/yolact_regnetx_800mf/predict`	POST	Boxes, confidences, mask areas (JSON)
`/api/video_feed`	GET	MJPEG preview stream with masks
`/api/config`	GET / POST	Confidence / NMS IOU thresholds

Implementation Notes

CPU post-processing (meta_arch=yolact): no on-chip NMS. The Detect pipeline (49,104 anchors → SSD decode → per-instance argmax → Fast NMS, top-200/class, max 100 detections) is a numpy port of the Model Zoo's instance_segmentation_postprocessing.py.
Candidate masks: sigmoid(proto @ coeffs^T) cropped to each detection box; visualized with mask_thresh=0.5 (YAML mask_threshold).
normalize_in_net ImageNet RGB; the app feeds raw uint8 RGB.
Preprocessing: plain bilinear resize to 512x512 (no letterbox, no pad).

Development Notes

Source module: src/rpi5_hailo8_yolact_regnetx_800mf/
Dockerfile: docker/hailo8/yolact_regnetx_800mf.dockerfile
Container: ghcr.io/seeed-projects/recomputer-hailo8-cv/yolact_regnetx_800mf:latest
Family: YOLACT (variants: regnetx_800mf / regnetx_1.6gf)

Invoer en uitvoer

Input: image, video, or USB camera frame. Output: COCO 80-class boxes with confidences, per-object mask pixel areas, and an MJPEG preview with colored instance masks.