EID-GAN: Generative Adversarial Nets for Extremely Imbalanced Data Augmentation

Efficiently generating sentence-level textual adversarial examples with Seq2seq Stacked Auto-Encoder

“…FEGR [55] learns to intrinsically decompose the driving scene for applications such as relighting. Lift3D [29] use NeRF to generate new objects and augment them to driving datasets, demonstrating the capability of NeRF to improve downstream task performance. The driving scene simulation provides a perfect test bed to evaluate the effectiveness of self-driving cars.…”

“…To address the above challenges and generate 3D adversarial examples in driving scenarios, we build Adv3D upon recent advances in NeRF [38] that provide both differentiable rendering and realistic synthesis. In order to generate physically realizable attacks, we model Adv3D in a patch-attack [44] manner and use an optimization-based approach that starts with a realistic NeRF object [29] to learn its 3D adversarial texture. We optimize the adversarial texture to minimize the predicted confidence of all objects in the scenes, while keeping shape unchanged.…”

“…(1) Note that we omit the direction term as suggested by [17]. The final predicted color of each pixel C(r) is computed by approximating the volume rendering integral using numerical quadrature [37]: We build our NeRF upon Lift3D [29]. Lift3D is a 3D generation framework that generates photorealistic objects by fitting multi-view images synthesized by 2D generative modes [23] using NeRF.…”

“…To achieve this, we add a semantic branch to Lift3d [29] to predict semantic part labels of the sampled points. We re-train Lift3d by fitting multi-view images and semantic labels generated by editGAN [31].…”

A Survey of Defect Detection Applications Based on Generative Adversarial Networks

“…ATTECGAN has been tested on KolektorSDD [96] and DAGM2007 [90] datasets, and its accuracy is 98.53% and 99.57%, respectively, with only a small number of samples. The literature [97,98,99,100,101] all proposed fabric defect detection methods based on GAN, and achieved relatively excellent detection performance. Table 4 shows the research application and performance of GAN based defect detection in the field of textile quality inspection.…”