본문 바로가기

AI

 (105)
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [1024, 1024]], which is output 0 of AsStridedBackward0, is at version 2; expected version 1 instead. 에러코드 전체 ''' RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [1024, 1024]], which is output 0 of AsStridedBackward0, is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True) ''' 구현하고자 했던 git..
[DAE-Former 논문 리뷰] - DAE-Former: Dual Attention-guided Efficient Transformer for Medical Image Segmentation *DAE-Former를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! DAE-Former paper: [2212.13504] DAE-Former: Dual Attention-guided Efficient Transformer for Medical Image Segmentation (arxiv.org) DAE-Former: Dual Attention-guided Efficient Transformer for Medical Image Segmentation Transformers have recently gained attention in the computer vision domain due to their ability to model long-range dependencies. Howev..
[GPT-1 논문 리뷰] - Improving Language Understanding by Generative Pre-Training *GPT-1를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! (학기중이라 블로그를 자주 못 쓰는데.. 나중에 시간되면 ChatGPT도 정리해서 올릴께요. 일단 간단한 GPT부터..ㅎㅎ) GPT-1 paper: https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf Contents 1. Simple Introduction 2. Background Knowledge: Transformer 3. Method - Unsupervised Stage - Supervised Stage 4. Result Simple Introduction 최근에 ..
[MCCNet 논문 리뷰] - Arbitrary Video Style Transfer via Multi-Channel Correlation *MCCNet를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! MCCNet paper: [2009.08003] Arbitrary Video Style Transfer via Multi-Channel Correlation (arxiv.org) Arbitrary Video Style Transfer via Multi-Channel Correlation Video style transfer is getting more attention in AI community for its numerous applications such as augmented reality and animation productions. Compared with traditional image style transfer, ..
[DietNeRF 논문 리뷰] - Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis *DietNeRF를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! DietNeRF paper: [2104.00677] Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis (arxiv.org) Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis We present DietNeRF, a 3D neural scene representation estimated from a few images. Neural Radiance Fields (NeRF) learn a continuous volumetric representation of a scene..
[MPS-Net 논문 리뷰] - Capturing Humans in Motion: Temporal-Attentive 3D Human Pose and Shape Estimation from Monocular Video *MPS-Net를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! MPS-Net project page: MPS-Net MPS-Net References [6] Hongsuk Choi, Gyeongsik Moon, Ju Yong Chang, and Kyoung Mu Lee. Beyond static features for temporally consistent 3D human pose and shape from a video. CVPR, 2021. [8] Carl Doersch and Andrew Zisserman. Sim2real transfer learning for 3D human mps-net.github.io MPS-Net github: GitHub - MPS-Net/MPS-Net_..
RuntimeError: view size is not compatible with input tensor's size and stride (at least one dimension spans across two contiguous subspaces). Use .reshape(...) instead. (Smplx 실행 에러) 갑자기 .view() function에 에러가 있다고 발생했다. 이 문제를 어떻게 해결할까? # smplx/lbs.py # Error code (below) transforms_mat = transform_mat( rot_mats.view(-1, 3, 3), rel_joints.view(-1, 3, 1)).view(-1, joints.shape[1], 4, 4) 위의 코드 block에서 오류가 일어난 상황이다. # smplx/lbs.py # Change code (below) transforms_mat = transform_mat( rot_mats.view(-1, 3, 3), rel_joints.contiguous().view(-1, 3, 1)).view(-1, joints.shape[1], 4, 4) ..
[wget 설치] - FileNotFoundError: [WinError 2] 지정된 파일을 찾을 수 없습니다(2). (MPS-Net 실행 에러) FFMPEG를 깔아서 실행되나 싶었더니, 갑자기 또 해당 에러가 발생했다. 후.. 이번에는 무엇이 문제일까? 원인 찾기 - 일단 terminal error를 통해서 multi_person_traker에 있는 mpt.py를 들어가보자. - Terminal error를 보니 여기 있는 YOLOv3를 만들 때 문제가 있는 것 같다. - 그러면 terminal에 경로를 따라서, yolov3/yolo.py에 들어가보자. - Terminal에서 출력된 downloading url 링크를 찾아내었다. - 해당 function을 보니, cmd에서 wget이라는 것을 실행시킬려고 한다. - ffmpeg처럼 wget도 설치를 해줘야 할 것이라고 느낌이 들었다. wget 설치 https://eternallybored.org..
[ffmpeg 설치] - FileNotFoundError: [WinError 2] 지정된 파일을 찾을 수 없습니다(1). (MPS-Net 실행 에러) 아무리 살펴봐도 경로도 다 맞는데, 어디서 파일을 찾을 수 없는지 이해할 수 없었던 찰나, ffmpeg가 매우 거슬렸다. FFMPEG 설치 https://ffmpeg.org/download.html#build-windows Download FFmpeg If you find FFmpeg useful, you are welcome to contribute by donating. More downloading options Git Repositories Since FFmpeg is developed with Git, multiple repositories from developers and groups of developers are available. Release Verification All FFmpe ..
Python argparse action='store_true'의 의미 # NeRF code def config_parser(): import argparse parser = argparse.ArgumentParser() parser.add_argument("--render_only", action='store_true', help='do not optimize, reload weights and render out render_poses path') parser.add_argument("--render_test", action='store_true', help='render the test set instead of render_poses path') return parser 요즘 github에 보면 argparse를 이용해서 인자를 받고 training 시키는 형태는 수..
[StylizedNeRF 논문 리뷰] - StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D Mutual Learning *StylizedNeRF를 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! StylizedNeRF project page: StylizedNeRF (geometrylearning.com) StylizedNeRF StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D Mutual Learning Yi-Hua Huang1,2 Yue He1,2 Yu-Jie Yuan1,2 Yu-Kun Lai3 Lin Gao1,2† --> † Corresponding author 1 Institute of Computing Technology, Chinese Ac geometrylearning.com StylizedNeRF github: Gi..
[AdaIN 논문 리뷰] - Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization *AdaIN을 위한 논문 리뷰 글입니다! 궁금하신 점은 댓글로 남겨주세요! AdaIN paper: [1703.06868] Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization (arxiv.org) Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization Gatys et al. recently introduced a neural algorithm that renders a content image in the style of another image, achieving so-called style transfer. However, their frame..
반응형

티스토리툴바