Localizing Adverts in Outdoor Scenes

Abstract

Online videos have witnessed an unprecedented growth over the last decade, owing to wide range of content creation. This provides the advertisement and marketing agencies plethora of opportunities for targeted advertisements. Such techniques involve replacing an existing advertisement in a video frame, with a new advertisement. However, such post-processing of online videos is mostly done manually by video editors. This is cumbersome and time-consuming. In this paper, we propose DeepAds -a deep neural network, based on the simple encoder-decoder architecture, that can accurately localize the position of an advert in a video frame. Our approach of localizing billboards in outdoor scenes using neural nets, is the first of its kind, and achieves the best performance. We benchmark our proposed method with other semantic segmentation algorithms, on a public dataset of outdoor scenes with manually annotated billboard binary maps.

Using U-Net network for efficient brain tumor segmentation in MRI images

“…This is particularly useful where the input data is the form of images. This architecture has several applications ranging from consumer videos [38,25], earth observations [39] and medical imaging [40,41,14].…”

An Advert Creation System for 3D Product Placements

“…Using large-scale annotated datasets of billboards [11,10], it provides them with an end-to-end framework for video editors to perform 2D advert placements. Their system assists in detecting frames in a video that contains a billboard [15], localizes the billboard in the detected frame [12], and subsequently replace the existing billboard with a new 2D advertisement. In this paper, we generalize this problem of product placements into any user-generated videos, and artificially augment 3D adverts into the existing scenes.…”

SYGNet: A SVD-YOLO based GhostNet for Real-time Driving Scene Parsing

“…Autonomous driving technology is the kind of core technology applied to the in-vehicle Artificial Intelligence (AI) field, and also can be defined as the key to realize smart cars, smart transportation and smart cities. Autonomous driving field consists of scenario generation [1,2], scene recognition [3,4], navigation [5], simulation [6], scene parsing [7][8][9], prediction [10,11] and motion estimation [12,13]. Among them, scene parsing is a fundamental core part and is used to obtain information of the vehicle itself and the surrounding environment, including vehicles, pedestrians, traffic signs, obstacles, through various sensors.…”