반응형

2024/03 4

[PaperReview] Seeing Through Fog Without Seeing Fog: Deep Multimodal Sensor Fusion in Unseen Adverse Weather

Seeing Through Fog Without Seeing Fog: Deep Multimodal Sensor Fusion in Unseen Adverse Weather Introduction & Related Works The fusion of multimodal sensor streams, such as camera, lidar, and radar measurements, plays a critical role in object detection for autonomous vehicles, which base their decision making on these inputs. While existing methods exploit redundant information in good environm..

AI/Paper Review 2024.03.31

[PaperReview] CycleGAN: Unpaired Image-to-Image Translationusing Cycle-Consistent Adversarial Networks

CycleGAN: Unpaired Image-to-Image Translationusing Cycle-Consistent Adversarial Networks Introduction & Related Works Image-to-image translation is a class of vision and graphics problems where the goal is to learn the mapping between an input image and an output image using a training set of aligned image pairs. Image-to-image translation은 말 그대로 pair image 간의 transformation function을 만드는 task이다..

AI/Paper Review 2024.03.31

[PaperReview] PatchmatchNet: Learned Multi-View Patchmatch Stereo

PatchmatchNet: Learned Multi-View Patchmatch Stereo Introduction Multiview stereo 방법론 중 Patchmatch를 더 빠른 속도로 수행하기 위한 방법이다. occlusion, illumination changes, untextured areas, non-Lambertian surfaces등에 의해 난이도가 높은 문제이다. 현재의 방법들은 대체로 3D cost volume을 만들고, 이를 3D convolution으로 해결하려고 한다. 기존 방법 : R-MVSNet [43] decouples the memory requirements from the depth range and sequentially processes the cost volu..

AI/Paper Review 2024.03.30

[PaperReview] UniSeg: A Unified Multi-Modal LiDAR Segmentation Networkand the OpenPCSeg Codebase

UniSeg: A Unified Multi-Modal LiDAR Segmentation Networkand the OpenPCSeg Codebase Introduction 색이나 texture 표현이 부족한 point cloud data를 위해서 RGB 이미지를 기용하고자 함 Sensor fusion을 위해 Learnable cross-Modal Association (LMA) module, Learnable cross-View Association module (LVA)을 제안 Pointcloud processing은 크게 세가지 형식으로 나뉜다. Point-view : pointcloud를 그대로 사용. 하지만 느린 연산과 neighboring point의 data를 잘 활용하지는 못하게 된다. Vo..

AI/Paper Review 2024.03.30
반응형