Topic
Sketch recognition
About: Sketch recognition is a research topic. Over the lifetime, 1611 publications have been published within this topic receiving 40284 citations.
Papers published on a yearly basis
Papers
More filters
TL;DR: Neural network and human psychophysics data on face recognition are presented which support the notion that recognition learning can be affected by the order in which images appear, as well as their spatial similarity.
Abstract: The view based approach to object recognition relies upon the co-activation of 2-D pictorial elements or features. This approach is limited to generalising recognition across transformations of objects in which considerable physical similarity is present in the stored 2-D images to which the object is being compared. It is, therefore, unclear how completely novel views of objects might correctly be assigned to known views of an object so as to allow correct recognition from any viewpoint. The answer to this problem may lie in the fact that in the real world we are presented with a further cue as to how we should associate these images, namely that we tend to view objects over extended periods of time. In this paper, neural network and human psychophysics data on face recognition are presented which support the notion that recognition learning can be affected by the order in which images appear, as well as their spatial similarity.
16 citations
01 Oct 2018
TL;DR: This paper proposes a novel point-based network with a compact architecture, named SketchPointNet, for robust sketch recognition, which achieves comparable performance on the challenging TU-Berlin dataset while it significantly reduces the network size.
Abstract: Sketch recognition is a challenging image processing task. In this paper, we propose a novel point-based network with a compact architecture, named SketchPointNet, for robust sketch recognition. Sketch features are hierarchically learned from three mini PointNets, by successively sampling and grouping 2D points in a bottom-up fashion. SketchPointNet exploits both temporal and spatial context in strokes during point sampling and grouping. By directly consuming the sparse points, SketchPointN et is very compact and efficient. Compared with state-of-the-art techniques, SketchPointNet achieves comparable performance on the challenging TU-Berlin dataset while it significantly reduces the network size.
16 citations
TL;DR: A neural network training protocol that consists of a large pool of unlabeled, synthetic samples generated from a small set of existing, labeled training samples, which leads to a significant error reduction compared to the baseline approaches is presented.
16 citations
TL;DR: Peng et al. as mentioned in this paper proposed a graph neural network (GNN) for learning representations of sketches from multiple graphs, which simultaneously capture global and local geometric stroke structures as well as temporal information.
Abstract: Learning meaningful representations of free-hand sketches remains a challenging task given the signal sparsity and the high-level abstraction of sketches. Existing techniques have focused on exploiting either the static nature of sketches with convolutional neural networks (CNNs) or the temporal sequential property with recurrent neural networks (RNNs). In this work, we propose a new representation of sketches as multiple sparsely connected graphs. We design a novel graph neural network (GNN), the multigraph transformer (MGT), for learning representations of sketches from multiple graphs, which simultaneously capture global and local geometric stroke structures as well as temporal information. We report extensive numerical experiments on a sketch recognition task to demonstrate the performance of the proposed approach. Particularly, MGT applied on 414k sketches from Google QuickDraw: 1) achieves a small recognition gap to the CNN-based performance upper bound (72.80% versus 74.22%) and infers faster than the CNN competitors and 2) outperforms all RNN-based models by a significant margin. To the best of our knowledge, this is the first work proposing to represent sketches as graphs and apply GNNs for sketch recognition. Code and trained models are available at https://github.com/PengBoXiangShang/multigraph_transformer.
16 citations
01 Oct 2006
TL;DR: A multimodal interface for sketch interpretation that relies on a multi-agent architecture and the design of the interpretation engine and the different agents are based on a user-centered approach where efficiency measure is defined as user satisfaction.
Abstract: We present a multimodal interface for sketch interpretation that relies on a multi-agent architecture. The design of the interpretation engine and the different agents are based on a user-centered approach where efficiency measure is defined as user satisfaction. So far, several graphical agents have been implemented for recognizing basic graphical objects (e.g., lines, circles, etc) as well as more complex (e.g., hatches, stairs, captions, etc) in architectural design. Besides, vocal agents have been developed for recognizing spoken annotations (e.g., dimensions) and interface commands. Realistic evaluations with professional users have demonstrated the potential interest of the proposed system.
16 citations