jim800121chen
9793a2efc1
使用者要求：前端模型庫只保留有適配裝置的模型（KL520 4 個 + KL720 3 個）。
原本 models.json 有 15 筆：
- 8 筆 ONNX framework 的 demo 模型（yolov5-face-detection /
imagenet-classification / person-detection / vehicle-classification /
hand-gesture-recognition / coco-object-detection / face-mask-detection /
license-plate-detection）— 都沒有實體 .nef 檔，是 placeholder metadata
- 7 筆 NEF framework 的實體模型（每個都有 filePath 指向 data/nef/）
現在只保留 7 筆實體模型：
KL520（4 個）：
- kl520-yolov5-detection (yolov5 no-upsample 640x640)
- kl520-fcos-detection (fcos-drk53s 512x512)
- kl520-ssd-face-detection (ssd_fd_lm 320x240)
- kl520-tiny-yolov3 (tiny_yolo_v3 416x416)
KL720（3 個）：
- kl720-yolov5-detection (yolov5 no-upsample 640x640)
- kl720-resnet18-classification (resnet18 224x224)
- kl720-fcos-detection (fcos-drk53s 512x512)
server/data/models.json 是 runtime 讀取，三平台（macOS/Windows/Linux）
共用同一份，改一次三平台全部生效。
驗證：
- python3 json.load 解析正常，7 筆 entries
- 每筆 filePath 指向的 .nef 實體檔都存在於 server/data/nef/{kl520,kl720}/
- 檔案大小：1-13 MB，合計 ~64 MB
- macOS dmg 重 build 163MB OK
- Bundle 內 Contents/Resources/data/models.json 更新為 7 筆
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
164 lines
6.3 KiB
JSON
164 lines
6.3 KiB
JSON
[
|
|
{
|
|
"id": "kl520-yolov5-detection",
|
|
"name": "YOLOv5 Detection (KL520)",
|
|
"description": "YOLOv5 object detection model compiled for Kneron KL520. No upsample variant optimized for NPU inference at 640x640 resolution.",
|
|
"thumbnail": "/images/models/yolov5-det.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["general", "multi-object"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 640, "height": 640},
|
|
"modelSize": 7200000,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.80,
|
|
"latencyMs": 50,
|
|
"fps": 20,
|
|
"supportedHardware": ["KL520"],
|
|
"labels": ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light"],
|
|
"filePath": "data/nef/kl520/kl520_20005_yolov5-noupsample_w640h640.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl520-fcos-detection",
|
|
"name": "FCOS Detection (KL520)",
|
|
"description": "FCOS (Fully Convolutional One-Stage) object detection with DarkNet53s backbone, compiled for KL520. Anchor-free detection at 512x512.",
|
|
"thumbnail": "/images/models/fcos-det.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["general", "multi-object"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 512, "height": 512},
|
|
"modelSize": 8900000,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.78,
|
|
"latencyMs": 45,
|
|
"fps": 22,
|
|
"supportedHardware": ["KL520"],
|
|
"labels": ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light"],
|
|
"filePath": "data/nef/kl520/kl520_20004_fcos-drk53s_w512h512.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl520-ssd-face-detection",
|
|
"name": "SSD Face Detection (KL520)",
|
|
"description": "SSD-based face detection with landmark localization, compiled for KL520. Lightweight model suitable for face detection and alignment tasks.",
|
|
"thumbnail": "/images/models/ssd-face.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["face", "security"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 320, "height": 240},
|
|
"modelSize": 1000000,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.85,
|
|
"latencyMs": 10,
|
|
"fps": 100,
|
|
"supportedHardware": ["KL520"],
|
|
"labels": ["face"],
|
|
"filePath": "data/nef/kl520/kl520_ssd_fd_lm.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl520-tiny-yolov3",
|
|
"name": "Tiny YOLOv3 (KL520)",
|
|
"description": "Tiny YOLOv3 object detection model compiled for KL520. Compact and fast model for general-purpose multi-object detection on edge devices.",
|
|
"thumbnail": "/images/models/tiny-yolov3.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["general", "multi-object"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 416, "height": 416},
|
|
"modelSize": 9400000,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.75,
|
|
"latencyMs": 35,
|
|
"fps": 28,
|
|
"supportedHardware": ["KL520"],
|
|
"labels": ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light"],
|
|
"filePath": "data/nef/kl520/kl520_tiny_yolo_v3.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl720-yolov5-detection",
|
|
"name": "YOLOv5 Detection (KL720)",
|
|
"description": "YOLOv5 object detection model compiled for Kneron KL720. No upsample variant optimized for KL720 NPU inference at 640x640 resolution with USB 3.0 throughput.",
|
|
"thumbnail": "/images/models/yolov5-det.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["general", "multi-object"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 640, "height": 640},
|
|
"modelSize": 10168348,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.82,
|
|
"latencyMs": 30,
|
|
"fps": 33,
|
|
"supportedHardware": ["KL720"],
|
|
"labels": ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light"],
|
|
"filePath": "data/nef/kl720/kl720_20005_yolov5-noupsample_w640h640.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl720-resnet18-classification",
|
|
"name": "ImageNet Classification ResNet18 (KL720)",
|
|
"description": "ResNet18-based image classification compiled for KL720. Supports 1000 ImageNet categories with fast inference via USB 3.0.",
|
|
"thumbnail": "/images/models/imagenet-cls.png",
|
|
"taskType": "classification",
|
|
"categories": ["general", "image-classification"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 224, "height": 224},
|
|
"modelSize": 12826804,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.78,
|
|
"latencyMs": 10,
|
|
"fps": 100,
|
|
"supportedHardware": ["KL720"],
|
|
"labels": ["airplane", "automobile", "bird", "cat", "deer", "dog", "frog", "horse", "ship", "truck"],
|
|
"filePath": "data/nef/kl720/kl720_20001_resnet18_w224h224.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "MIT",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
},
|
|
{
|
|
"id": "kl720-fcos-detection",
|
|
"name": "FCOS Detection (KL720)",
|
|
"description": "FCOS (Fully Convolutional One-Stage) object detection with DarkNet53s backbone, compiled for KL720. Anchor-free detection at 512x512.",
|
|
"thumbnail": "/images/models/fcos-det.png",
|
|
"taskType": "object_detection",
|
|
"categories": ["general", "multi-object"],
|
|
"framework": "NEF",
|
|
"inputSize": {"width": 512, "height": 512},
|
|
"modelSize": 13004640,
|
|
"quantization": "INT8",
|
|
"accuracy": 0.80,
|
|
"latencyMs": 30,
|
|
"fps": 33,
|
|
"supportedHardware": ["KL720"],
|
|
"labels": ["person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light"],
|
|
"filePath": "data/nef/kl720/kl720_20004_fcos-drk53s_w512h512.nef",
|
|
"version": "1.0.0",
|
|
"author": "Kneron",
|
|
"license": "Apache-2.0",
|
|
"createdAt": "2024-01-01T00:00:00Z",
|
|
"updatedAt": "2024-01-01T00:00:00Z"
|
|
}
|
|
]
|