목록Deep Learning (12)

Shakerato

pytorch CUDNN_STATUS_EXECUTION_FAILED error (Anaconda3)

[previous setup]conda install pytorch cuda90 -c pytorch (installed)pytorch 0.4.0 py36_cuda90_cudnn7he774522_1 [cuda90] pytorch --> If you are using the latest GPU in the above environment, for example, 2080ti,you may get an error message when you run pytorch code using GPU as shown in below. CUDNN_STATUS_EXECUTION_FAILED --> [solution setup]1. install cuda 10.0 and new cudnn for this2. conda upd..
Research 2019. 2. 23. 20:37
Video Action Classification using 3D ResNet

Reference:https://github.com/kenshohara/video-classification-3d-cnn-pytorch (Linux)1. Install Anaconda and Setup Pytorch (GPU)2. Download code from Github (https://github.com/kenshohara/video-classification-3d-cnn-pytorch)3. Download Pre-trained Model and copy to source code folder (resnext-101-kinetics.pth) https://drive.google.com/drive/folders/1zvl89AgFAApbH0At-gMuZSeQB_LpNP-M4. Install FFmpe..
Research 2019. 2. 18. 12:13
convert *.mat data file to json file

(matlab이 설치되어 있어야 함) 1. 코드 다운로드 'https://github.com/leonid-pishchulin/poseval' 2. matlab 폴더에 임의의 폴더(예를들어 test)를 만들기 3. 변경하고자 하는 matlab 데이터 파일 복사(*.mat)을 위 폴더에 복사 4. starup.m 을 실행하여 addpath addpath('./external/jsonlab') 5. mat2json 를 위 경로를 포함하여 실행, json 형테로 모두 바뀌어 동일한 폴더에 저장됨 mat2json('./test')
Research 2019. 2. 14. 02:24
Theano (using GPU) install in Windows 10

1. Install Anaconda3, GPU driver, CUDA, cudnn2. Run 'cmd' as administrator3. conda create -n [myenv] python=3.6 anaconda4. activate [myenv]5. conda install theano pygpu(if you got error, try this command: "conda install -c conda-forge pygpu") 6. conda install numpy scipy mkl-service libpython 7. Create a file "C:\Users\[Username]\.theanorc"[global]floatX = float32device = cuda0 8. Test code on G..
Research 2019. 2. 7. 01:05
딥러닝 객체인식 네트워크 retinanet 내 데이터셋으로 학습하는 방법

참고: https://github.com/fizyr/keras-retinanet 1. 실행 환경 구축 1.1. Anaconda (python=3.6) 개발 환경 생성 conda create -n myenv python=3.6 1.2. Anaconda 개발 환경 활성화 (윈도우) activate myenv 1.3. 관리자 권한으로 cmd 실행 1.4. python setup.py install 1.5. pip install Keras==2.2.4 2. 학습할 데이터셋 만들기 2.1. 학습할 이미지 준비하기 2.2. 학습할 이미지에 있는 인식할 객체의 Bounding Box 위치 알아내기 2.3. 모든 학습할 이미지를 특정 폴더에 path/to/image.jpg 모두 저장 2.4. Annotations fo..
Research 2019. 1. 17. 00:56
jupyter notebook setup: an anaconda environment as a kernel

1. Install ipykernel each environmentsconda install ipykernel 2. Set jupyter notebook kernel in the anaconda default environmentType 'activate' in cmdpython -m ipykernel install --user --name --display-name "" 3. Run the juypter-notebookC:\ProgramData\Anaconda3\Scripts\jupyter-notebook.exe --ip 4. Change kernel Click *.ipynb - [Kernel tap] - [change kernel] - select a kernel
Research 2018. 7. 26. 06:38
Prev 1 2 Next