Hardware components | ||||||
 |
| × | 1 | |||
Software apps and online services | ||||||
 |
| |||||
| ||||||
| ||||||
| ||||||
The System Configuration we have used is as below:
Intel Core i9-13900KF 13th Gen. Processor
(24 Core, 32 Thread, 36MB Cache)
Motherboard Z790 Chipset
Corsair 64GB (2x32GB) DDR5 6000MHzWD 1 TB Blue NVME x 2
CPU Air Cooler Deepcool AG620Power Supply 850 Watt Cabinet Full Tower Antesports 410TG
Prior to the start of the implementation of Video Analytics, we installed all the AMD Radeon W7900 Pro GPU related libraries and Pytorch-Tensorflow frameworks and tested the workings of all the libraries.
N.B.:- Our Main Project Problem Statement was related to VLM (Visual Language Model) on customized medical dataset. But due to new entry of the AMD in GPU Market, there was some issues related to implementation of some of the necessary libraries which are still under development. Time Limitation to successfully complete that task was also an issue because of late shipment+delivery of the AMD Radeon W7900 Pro GPU.
So, here we going going to submit an other relatively simpler application solution. i.e. Video Analytics for Football Match.
We downloaded some already annotated FootBall Live Match related datasets from ROBOFLOW.
Dataset used: https://universe.roboflow.com/nikhil-chapre-xgndf/detect-players-dgxz0
We use the Ultralytics library, which maintains all the Yolo models:
pip install ultralyticsWe define the customized data related yaml file as below
train: /home/smss/datasets/football_yolov8/train/images
val: /home/smss/datasets/football_yolov8/valid/images
test: /home/smss/datasets/football_yolov8/test/images
nc: 3
names: ['Ball', 'Player', 'Refree']
roboflow:
workspace: nikhil-chapre-xgndf
project: detect-players-dgxz0
version: 7
license: CC BY 4.0
url: https://universe.roboflow.com/nikhil-chapre-xgndf/detect-players-dgxz0/dataset/7Next, we train the downloaded dataset using the CLI command.
yolo detect train data=data.yaml model=yolov8n.yaml epochs=100 imgsz=640The command-line output & batch training results are shown below:-
/home/smss/.local/lib/python3.10/site-packages/matplotlib/projections/__init__.py:63: UserWarning: Unable to import Axes3D. This may be due to multiple versions of Matplotlib being installed (e.g. as a system package and as a pip package). As a result, the 3D projection is not available.
warnings.warn("Unable to import Axes3D. This may be due to multiple versions of "
Ultralytics YOLOv8.2.70 🚀 Python-3.10.12 torch-2.3.1+rocm6.0 CUDA:0 (AMD Radeon PRO W7900, 46064MiB)
engine/trainer: task=detect, mode=train, model=yolov8n.yaml, data=data.yaml, epochs=100, time=None, patience=100, batch=16, imgsz=640, save=True, save_period=-1, cache=False, device=None, workers=8, project=None, name=train3, exist_ok=False, pretrained=True, optimizer=auto, verbose=True, seed=0, deterministic=True, single_cls=False, rect=False, cos_lr=False, close_mosaic=10, resume=False, amp=True, fraction=1.0, profile=False, freeze=None, multi_scale=False, overlap_mask=True, mask_ratio=4, dropout=0.0, val=True, split=val, save_json=False, save_hybrid=False, conf=None, iou=0.7, max_det=300, half=False, dnn=False, plots=True, source=None, vid_stride=1, stream_buffer=False, visualize=False, augment=False, agnostic_nms=False, classes=None, retina_masks=False, embed=None, show=False, save_frames=False, save_txt=False, save_conf=False, save_crop=False, show_labels=True, show_conf=True, show_boxes=True, line_width=None, format=torchscript, keras=False, optimize=False, int8=False, dynamic=False, simplify=False, opset=None, workspace=4, nms=False, lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=7.5, cls=0.5, dfl=1.5, pose=12.0, kobj=1.0, label_smoothing=0.0, nbs=64, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, bgr=0.0, mosaic=1.0, mixup=0.0, copy_paste=0.0, auto_augment=randaugment, erasing=0.4, crop_fraction=1.0, cfg=None, tracker=botsort.yaml, save_dir=runs/detect/train3
Overriding model.yaml nc=80 with nc=3
from n params module arguments
0 -1 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
1 -1 1 4672 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2]
2 -1 1 7360 ultralytics.nn.modules.block.C2f [32, 32, 1, True]
3 -1 1 18560 ultralytics.nn.modules.conv.Conv [32, 64, 3, 2]
4 -1 2 49664 ultralytics.nn.modules.block.C2f [64, 64, 2, True]
5 -1 1 73984 ultralytics.nn.modules.conv.Conv [64, 128, 3, 2]
6 -1 2 197632 ultralytics.nn.modules.block.C2f [128, 128, 2, True]
7 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
8 -1 1 460288 ultralytics.nn.modules.block.C2f [256, 256, 1, True]
9 -1 1 164608 ultralytics.nn.modules.block.SPPF [256, 256, 5]
10 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
11 [-1, 6] 1 0 ultralytics.nn.modules.conv.Concat [1]
12 -1 1 148224 ultralytics.nn.modules.block.C2f [384, 128, 1]
13 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
14 [-1, 4] 1 0 ultralytics.nn.modules.conv.Concat [1]
15 -1 1 37248 ultralytics.nn.modules.block.C2f [192, 64, 1]
16 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
17 [-1, 12] 1 0 ultralytics.nn.modules.conv.Concat [1]
18 -1 1 123648 ultralytics.nn.modules.block.C2f [192, 128, 1]
19 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
20 [-1, 9] 1 0 ultralytics.nn.modules.conv.Concat [1]
21 -1 1 493056 ultralytics.nn.modules.block.C2f [384, 256, 1]
22 [15, 18, 21] 1 751897 ultralytics.nn.modules.head.Detect [3, [64, 128, 256]]
YOLOv8n summary: 225 layers, 3,011,433 parameters, 3,011,417 gradients, 8.2 GFLOPs
TensorBoard: Start with 'tensorboard --logdir runs/detect/train3', view at http://localhost:6006/
Freezing layer 'model.22.dfl.conv.weight'
AMP: running Automatic Mixed Precision (AMP) checks with YOLOv8n...
AMP: checks passed ✅
train: Scanning /home/smss/datasets/football_yolov8/train/labels.cache... 1023 images, 4 backgrounds,
val: Scanning /home/smss/datasets/football_yolov8/valid/labels.cache... 293 images, 0 backgrounds, 0
Plotting labels to runs/detect/train3/labels.jpg...
optimizer: 'optimizer=auto' found, ignoring 'lr0=0.01' and 'momentum=0.937' and determining best 'optimizer', 'lr0' and 'momentum' automatically...
optimizer: AdamW(lr=0.001429, momentum=0.9) with parameter groups 57 weight(decay=0.0), 64 weight(decay=0.0005), 63 bias(decay=0.0)
TensorBoard: model graph visualization added ✅
Image sizes 640 train, 640 val
Using 8 dataloader workers
Logging results to runs/detect/train3
Starting training for 100 epochs...
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
1/100 0.388G 4.832 4.859 4.046 530 640: 100%|██████████| 64/64
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.000511 0.00269 0.00028 0.000106
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
2/100 2.58G 2.965 2.017 2.256 336 640: 100%|██████████| 64/64
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.797 0.101 0.0862 0.0258
........
........
........
........
........
........
........
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
98/100 2.17G 1.136 0.543 0.9401 224 640: 100%|██████████| 64/64
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.83 0.666 0.711 0.437
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
99/100 2.21G 1.136 0.5415 0.9413 180 640: 100%|██████████| 64/64
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.861 0.655 0.712 0.44
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
100/100 2.21G 1.13 0.5428 0.9424 199 640: 100%|██████████| 64/64
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.864 0.661 0.715 0.44
100 epochs completed in 0.186 hours.
Optimizer stripped from runs/detect/train3/weights/last.pt, 6.2MB
Optimizer stripped from runs/detect/train3/weights/best.pt, 6.2MB
Validating runs/detect/train3/weights/best.pt...
Ultralytics YOLOv8.2.70 🚀 Python-3.10.12 torch-2.3.1+rocm6.0 CUDA:0 (AMD Radeon PRO W7900, 46064MiB)
YOLOv8n summary (fused): 168 layers, 3,006,233 parameters, 0 gradients, 8.1 GFLOPs
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████
all 293 4381 0.865 0.662 0.715 0.439
Ball 214 215 0.883 0.293 0.346 0.148
Player 293 3811 0.924 0.954 0.974 0.65
Player 220 355 0.789 0.739 0.826 0.519
Speed: 0.0ms preprocess, 0.4ms inference, 0.0ms loss, 0.4ms postprocess per image
Results saved to runs/detect/train3
💡 Learn more at https://docs.ultralytics.com/modes/trainTrain Batch Results
1 / 6
After the training is done, we predict the player and ball detection using the trained best model from the custom dataset on a different downloaded non-trained type of video using the CLI command.
yolo detect predict model=/home/smss/datasets/football_yolov8/runs/detect/train3/weights/best.pt source=video.mp4 show=TrueThe command-line output & output prediction result video is shown below:-
/home/smss/.local/lib/python3.10/site-packages/matplotlib/projections/__init__.py:63: UserWarning: Unable to import Axes3D. This may be due to multiple versions of Matplotlib being installed (e.g. as a system package and as a pip package). As a result, the 3D projection is not available.
warnings.warn("Unable to import Axes3D. This may be due to multiple versions of "
Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
Ultralytics YOLOv8.2.70 🚀 Python-3.10.12 torch-2.3.1+rocm6.0 CUDA:0 (AMD Radeon PRO W7900, 46064MiB)
YOLOv8n summary (fused): 168 layers, 3,006,233 parameters, 0 gradients, 8.1 GFLOPs
video 1/1 (frame 1/258) /home/smss/datasets/football_yolov8/video.mp4: 384x640 8 Players, 1 Player, 5.3ms
video 1/1 (frame 2/258) /home/smss/datasets/football_yolov8/video.mp4: 384x640 7 Players, 1 Player, 2.8ms
........
........
........
........
........
........
........
video 1/1 (frame 256/258) /home/smss/datasets/football_yolov8/video.mp4: 384x640 10 Players, 3 Players, 2.8ms
video 1/1 (frame 257/258) /home/smss/datasets/football_yolov8/video.mp4: 384x640 10 Players, 3 Players, 2.7ms
video 1/1 (frame 258/258) /home/smss/datasets/football_yolov8/video.mp4: 384x640 11 Players, 4 Players, 2.7ms
Speed: 1.3ms preprocess, 2.8ms inference, 0.9ms postprocess per image at shape (1, 3, 384, 640)
Results saved to runs/detect/predict14
💡 Learn more at https://docs.ultralytics.com/modes/predictrocm-smi terminal command used to show the utilization details of the AMD Radeon Pro W7900 GPU. Initially the percentage usuage of GPU is 7% which increased to 99% drastically as soon as the train command run which supports that the training is supported via the AMD Radeon Pro W7900 GPU.
smss@smss:~/datasets/football_yolov8$ rocm-smi
========================================== ROCm System Management Interface ==========================================
==================================================== Concise Info ====================================================
Device Node IDs Temp Power Partitions SCLK MCLK Fan Perf PwrCap VRAM% GPU%
(DID, GUID) (Edge) (Avg) (Mem, Compute, ID)
======================================================================================================================
0 1 0x7448, 19246 33.0°C 30.0W N/A, N/A, 0 158Mhz 1124Mhz 20.78% auto 241.0W 30% 7%
======================================================================================================================
================================================ End of ROCm SMI Log =================================================
smss@smss:~/datasets/football_yolov8$ rocm-smi
=========================================== ROCm System Management Interface ===========================================
===================================================== Concise Info =====================================================
Device Node IDs Temp Power Partitions SCLK MCLK Fan Perf PwrCap VRAM% GPU%
(DID, GUID) (Edge) (Avg) (Mem, Compute, ID)
========================================================================================================================
0 1 0x7448, 19246 68.0°C 241.0W N/A, N/A, 0 2042Mhz 1124Mhz 48.63% auto 241.0W 30% 99%
========================================================================================================================
================================================= End of ROCm SMI Log ==================================================_6xIzVSSxO3.jpeg)
_6xIzVSSxO3.jpeg){kind=link}
Comments
Please log in or sign up to comment.