Error while generation detector results

2 次查看（过去 30 天）
Divya
Divya 2024-5-10
回答： Naren 2024-5-10
I am using Yolov4 with darknet-53-coco. I have trained the detector but unable to generate detector results using
detectionResults = detect(detector,testData,Threshold=0.01);
After running the code I am getting the following error
Error using dlnetwork/predict
Execution failed during layer(s) 'conv_2'.
Error in yolov4ObjectDetector>iPredictActivations (line 1223)
[features{:}] = predict(network, dlX, 'Acceleration',acceleration);
Error in yolov4ObjectDetector/predict (line 555)
features = iPredictActivations(network, imgBatch, anchorBoxes, predictParams.Acceleration, isRotatedBox);
Error in vision.internal.detector.ObjectDetector/performDetect (line 72)
features = this.predict(Ipreprocessed, params);
Error in yolov4ObjectDetector/detect (line 441)
[varargout{1:nargout}] = performDetect(detector, I, params);
Caused by:
Error using builtin
Out of memory.

请先登录，再回答此问题。

回答（1 个）

Naren
Naren 2024-5-10
Hello,
The issue seems to be due to low GPU memory. I suggest reducing the minibatch size, which could be effective. If the error still occurs even after reducing the minibatch size, another possible solution is to change the network to "tiny-yolov4ObjectDetector" instead of "csp-darknet53-coco". Apart from these options you can also try reducing the size of network input image.
Regards.

类别

在 Help Center 和 File Exchange 中查找有关 Deep Learning Toolbox 的更多信息

标签

产品


版本

R2024a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by