🌟 YoloV8s Dataset v0.2

Metadata

• Type: Yolov8s
• Filter: [64,128,256,512,768]
• Kernel Size: 3
• Train Size: 1400 × 1400
• Images Size: 963 × 1400

Preprocessing Images

• Auto-Orient: Applied
• Resize: Fit (white edges) in 963x1400
• Auto-Adjust Contrast: Using Contrast Stretching

Augmentations for Images

• Flip: Horizontal, Vertical
• Rotation: Between -15° and +15°
• Grayscale: Apply to 100% of images
• Exposure: Between -10% and +10%
• Blur: Up to 1px
• Noise: Up to 0.5% of pixels

Training data

• 480 Images (with generated images, original size 200)
• 420 for training (image set)
• 40 for validation (image set)
• 20 for test (image set)
• 4430 annotations (1846 without the generated images)

Train Accuracy

Precision(M)	Recall(M)	mAP50(M)	mAP50-95(M)	CLS Loss	DFL Loss
0.96524	0.97068	0.98096	0.91488	0.31534	1.04077

✨ YoloV11s Dataset v0.2

Metadata

• Type: Yolo
• Filter: [64, 128, 256, 512, 1024]
• Kernel Size: 3
• Train Size: 1400 × 1400
• Images Size: 963 × 1400

Preprocessing Images

• Auto-Orient: Applied
• Resize: Fit (white edges) in 963x1400
• Auto-Adjust Contrast: Using Contrast Stretching

Augmentations for Images

• Flip: Horizontal, Vertical
• Rotation: Between -15° and +15°
• Grayscale: Apply to 100% of images
• Exposure: Between -10% and +10%
• Blur: Up to 1px
• Noise: Up to 0.5% of pixels

Training data

• 480 Images (with generated images, original size 200)
• 420 for training (image set)
• 40 for validation (image set)
• 20 for test (image set)
• 4430 annotations (1846 without the generated images)

Train Accuracy

Precision(M)	Recall(M)	mAP50(M)	mAP50-95(M)	CLS Loss	DFL Loss
0.96811	0.96575	0.98058	0.92454	0.27078	1.03036

⭐ YoloV8n Dataset v0.1

Metadata

• Type: Yolo
• Seed: 369
• Filter: [64, 128, 256, 512, 768]
• Kernel Size: 3
• Dropout: 0.0
• Train Size: 1280 × 1280
• Images Size: 963 × 1400

Preprocessing Images

• Auto-Orient: Applied
• Resize: Stretch to 963 × 1400
• Auto-Adjust Contrast: Using Contrast Stretching
• Grayscale: Applied

Augmentations for Images

• Outputs per training example: 3
• Flip: Horizontal, Vertical
• Blur: Up to 0.5px
• Noise: Up to 0.5% of pixels

Training data

• 283 Images (with generated images, original size 117)
• 249 for training (image set)
• 22 for validation (image set)
• 12 for test (image set)

Train Accuracy

Precision	Recall	mAP50	mAP50-95	CLS Loss	DFL Loss	Fitness
0.96808	0.9731	0.97648	0.93864	0.25748	0.98696	1.87714

☢️ [Beta] Model - 473

Metadata

• ID: 473
• Type: Unet
• Val Accuracy: 0.7444
• filter: [32, 64, 128, 256, 512]
• Kernel Size: 3
• Dropout: 0.2
• Size image: 512 x 768

Training data

• 3.882 Images (with generated images, original size 1.941)
• 1552 for training (image set)
• 388 for validation (image set)

Kaette Kudasai! Akutsu-san

• Ch. 2 - 49

The Dangers in My Heart

• Ch. 1 - 10

Yofukashi no Uta

• Ch. 1 - 10
• Ch.190 - 196

❌ [Beta] Model - 382

Metadata

• ID: 382
• Type: Unet
• Val Accuracy: [0.90806] <-- That's a mistake, it's not real, it's not possible
• filter: [32, 64, 128, 256, 512]
• Kernel Size: 7
• Dropout: 0.2
• Size image: 512 x 768

Training data

• 1.956 Images (with generated images, original size 978)
• 783 for training (image set)
• 195 for validation (image set)

Oroka na Tenshi wa Akuma to Odoru

• Ch. 1

The Dangers in My Heart

• Ch. 1 - 10

Yofukashi no Uta

• Ch. 1 - 10
• Ch.190 - 196