Overall framework:MPNN、GN、NLNN 整体结构:模型的输入为二进制代码的控制流图,模型的整体结构如图所示,包含semantic-aware 模块、structural-aware模块、order-aware模块。在semantic-aware模块,模型将控制流图作为输入,使用BERT对token embedding作预训练,得到block embedding。在structural-aware模块,使用MPNN算法得到graph ...
Deep learning; object detection; semantic features Abstract Recent deep convolutional neural network-based object detectors have shown promising performance when detecting large objects, but they are still limited in detecting small or partially occluded ones—in part because such objects convey limited inf...
Object detection results for infrared, visible and fused images from the MFNet dataset. The YOLOv5 detector, pre-trained on the Coco dataset is deployed to achieve object detection. If this work is helpful to you, please cite it as:
Improving semantic video retrieval models by training with a relevance-aware online mining strategy a general training strategy overcoming these two limitations.We implement RANP into two novel loss functions, Triplet-RANP and NCE-RANP.Using RANP, we ... A Falcon,G Serra,O Lanz - 《Computer ...
We propose an object detection system that relies on a multi-region deep convolutional neural network (CNN) that also encodes semantic segmentation-aware features. The resulting CNN-based representation aims at capturing a diverse set of discriminative appearance factors and exhibits localizatio...
从FCN开始,最初的困难应该来自如果让CNN能区分不同位置的物体,这一方面衍生除了很多办法,比如COCO 2015里MSRA那篇用了一个Multi-task的方法,COCO 2016MSRA那篇又用了position-sensitive的map来做到instance-aware。从稍难的segmentation的问题到实质上是detection的问题还是有些跨度的。 发布于 2017-03-26 20:42 ...
SA-BEV: Generating Semantic-Aware Bird's-Eye-View Feature for Multi-view 3D Object Detection Supplementary Material This supplementary material provides more implementa- tion details on SA-BEV in Sec. A, more experiments results in Sec. B and additional visualizations in Sec. C...
Task: Binary code similarity detection 传统: graph matching algo - 缺点: slow & inaccurate 新法(本文所属): control-flow graph + (人工)筛选的特征 + GNN-> graph embedding 本文: 1. semantic-aware neural network 2. BERT预训练: 1个token-level, 1个block-level,2个graph level ...
Open-Vocabulary Object Detection Universal Semantic Segmentation [Semantic-SAM]|ECCV'24| Semantic-SAM: Segment and Recognize Anything at Any Granularity |[pdf]|[code] [Open-Vocabulary SAM]|ECCV'24| Open-Vocabulary SAM: Segment and Recognize Twenty-thousand Classes Interactively |[pdf]|[code] ...
摘要:本文主要是给出一个类别感知的语义边缘检测算法。传统的边缘检测本身就是一个具有挑战性的二元问题,相比之下类别感知的语义边缘检测是一个更具有挑战性的多元问题。因为边缘像素出现在属于两个或更多个语义类的轮廓或连接点中,所以本文对每个边缘像素与至少两个类别相关联这个问题进行建模,并提出了一种新的基于Re...