Skip to main content

Posts

Showing posts from 2022

Detection and Segmentation

 Add layers during training Net to Net [?] Ian Goodfellow Network Morphism Microsoft Deep Visual-Semantic: Alignments for Generating Image Descriptions Types of RNN - Vanilla RNN / Simple RNN / Elman RNN - Long Short Term Memory (LSTM) - Helps improve gradient flow during back prop Computer Vision Tasks  - Classification [already done] - Very basic  - Localization  - Segmentation  - Detection 1. Semantic Segmentation - Grass, Cat, Tree, Sky in 1 image. [No object just pixels] - No box but all pixels are classified. 2. Classification + Localization - Class = Cat and a box around cat [Single Object] 3. Object detection  - Box around objects - Dog, Dog, Cat 4. Instance Segmentation - All pixels are classified - Dog Cat Dog  3 and 4 - Multiple objects. Semantic Segmentation  - For every pixel we want to say what it is - cat, grass, sky, etc - Two Cows will be classified as cow, both are not differentiated like cow 1 and cow 2. - Instance segmentation does this. - Just labeling the

Initial Apps HP

 

Installing Nvidia Driver and CUDA Ubuntu Victus

  Get corresponding CUDA for your TF Version: https://www.tensorflow.org/install/source#gpu https://developer.nvidia.com/cuda-11-8-0-download-archive?target_os=Linux&target_arch=x86_64&Distribution=Ubuntu&target_version=22.04&target_type=runfile_local Example: wget https://developer.download.nvidia.com/compute/cuda/11.8.0/local_installers/cuda_11.8.0_520.61.05_linux.run sudo sh cuda_11.2.2_460.32.03_linux.run --override --override to override gcc version check [Existing package manage installation of driver found] : Continue Update: I only installed Toolkit no driver/samples/demo suite/doc - note driver version is 520 export PATH=/usr/local/cuda-11.8/bin${PATH:+:${PATH}} export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}} export CUDA_HOME=/usr/local/cuda Now install CUDNN: https://developer.nvidia.com/rdp/cudnn-archive For my case: Download cuDNN v8.6.0 (October 3rd, 2022), for CUDA 11.x Download tar: [doesn't work anymore withou

Install WSL2 Ubuntu

wsl enable from windows features wsl enter in command prompt WSL 2 requires an update to its kernel component. For information please visit https://aka.ms/wsl2kernel dism.exe /online /enable-feature /featurename:VirtualMachinePlatform /all /norestart wsl --set-default-version 2 Still got: WSL 2 requires an update to its kernel component. For information please visit https://aka.ms/wsl2kernel wsl --update Failed to attach disk 'X:\WSL_IMPORT\ubuntu\ext4.vhdx' to WSL2: The system cannot find the path specified. Error code: Wsl/Service/CreateInstance/MountVhd/HCS/ERROR_PATH_NOT_FOUND wsl --unregister ubuntu wsl --install sudo apt install gcc  [Didn't help but installed first] sudo apt install g++ [Helped] - for detectron2 ImportError: libGL.so.1: cannot open shared object file: No such file or directory sudo apt-get install libgl1 import tensorflow as tf print(len(tf.config.list_physical_devices('GPU'))) from tensorflow.python.client import device_lib print(device_lib.