Abstract:Sketch-based 3D shape retrieval is a challenging task due to the large domain discrepancy between sketches and 3D shapes. Since existing methods are trained and evaluated on the same categories, they cannot effectively recognize the categories that have not been used during training. In this paper, we propose a novel domain disentangled generative adversarial network (DD-GAN) for zero-shot sketch-based 3D retrieval, which can retrieve the unseen categories that are not accessed during training. Specifically, we first generate domain-invariant features and domain-specific features by disentangling the learned features of sketches and 3D shapes, where the domain-invariant features are used to align with the corresponding word embeddings. Then, we develop a generative adversarial network that combines the domain-specific features of the seen categories with the aligned domain-invariant features to synthesize samples, where the synthesized samples of the unseen categories are generated by using the corresponding word embeddings. Finally, we use the synthesized samples of the unseen categories combined with the real samples of the seen categories to train the network for retrieval, so that the unseen categories can be recognized. In order to reduce the domain shift problem, we utilize unlabeled unseen samples to enhance the discrimination ability of the discriminator. With the discriminator distinguishing the generated samples from the unlabeled unseen samples, the generator can generate more realistic unseen samples. Extensive experiments on the SHREC'13 and SHREC'14 datasets show that our method significantly improves the retrieval performance of the unseen categories.

Adversarial Training For Sketch Retrieval

SceneSketcher-v2: Fine-Grained Scene-Level Sketch-Based Image Retrieval Using Adaptive GCNs

Teaching GANs to sketch in vector format

GGD-GAN: Gradient-Guided Dual-Branch Adversarial Networks for Relic Sketch Generation

Sketch Your Own GAN

Domain Disentangled Generative Adversarial Network for Zero-Shot Sketch-Based 3D Shape Retrieval.

SketchGAN: Joint Sketch Completion and Recognition with Generative Adversarial Network.

Picture that Sketch: Photorealistic Image Generation from Abstract Sketches

SmartPaint: a Co-Creative Drawing System Based on Generative Adversarial Networks

How to Handle Sketch-Abstraction in Sketch-Based Image Retrieval?

Cross-domain representation learning by domain-migration generative adversarial network for sketch based image retrieval

Staged Sketch-to-Image Synthesis via Semi-supervised Generative Adversarial Networks

Creative Sketch Generation

Towards Criminal Sketching with Generative Adversarial Network

Adversarial Open Domain Adaption Framework (AODA): Sketch-to-Photo Synthesis

Sketch Less for More: On-the-Fly Fine-Grained Sketch Based Image Retrieval

R(2)GAN: Cross-modal Recipe Retrieval with Generative Adversarial Network

Unsupervised Adversarial Attacks on Deep Feature-based Retrieval with GAN

Instance Image Retrieval with Generative Adversarial Training

Sketch to Building: Architecture Image Translation Based on GAN

On Learning Semantic Representations for Large-Scale Abstract Sketches