No abstract
Generative Adversarial Networks (GANs) have gathered a lot of attention from the computer vision community, yielding impressive results for image generation. Advances in the adversarial generation of natural language from noise however are not commensurate with the progress made in generating images, and still lag far behind likelihood based methods. In this paper, we take a step towards generating natural language with a GAN objective alone. We introduce a simple baseline that addresses the discrete output space problem without relying on gradient estimators and show that it is able to achieve state-of-the-art results on a Chinese poem generation dataset. We present quantitative results on generating sentences from context-free and probabilistic context-free grammars, and qualitative language modeling results. A conditional version is also described that can generate sequences conditioned on sentence characteristics.
This paper presents a Deep Bidirectional Long Short Term Memory (LSTM) based Recurrent Neural Network architecture for text recognition. This architecture uses Connectionist Temporal Classification (CTC) for training to learn the labels of an unsegmented sequence with unknown alignment. This work is motivated by the results of Deep Neural Networks for isolated numeral recognition and improved speech recognition using Deep BLSTM based approaches. Deep BLSTM architecture is chosen due to its ability to access long range context, learn sequence alignment and work without the need of segmented data. Due to the use of CTC and forward backward algorithms for alignment of output labels, there are no unicode re-ordering issues, thus no need of lexicon or postprocessing schemes. This is a script independent and segmentation free approach. This system has been implemented for the recognition of unsegmented words of printed Oriya text. This system achieves 4.18% character level error and 12.11% word error rate on printed Oriya text.
We infer and generate three-dimensional (3D) scene information from a single input image and without supervision. This problem is under-explored, with most prior work relying on supervision from, e.g., 3D groundtruth, multiple images of a scene, image silhouettes or key-points. We propose Pix2Shape, an approach to solve this problem with four component: (i) an encoder that infers the latent 3D representation from an image, (ii) a decoder that generates an explicit 2.5D surfelbased reconstruction of a scene -from the latent code -(iii) a differentiable renderer that synthesizes a 2D image from the surfel representation, and (iv) a critic network trained to discriminate between images generated by the decoder-renderer and those from a training distribution. Pix2Shape can generate complex 3D scenes that scale with the view-dependent on-screen resolution, unlike representations that capture world-space resolution, i.e., voxels or meshes. We show that Pix2Shape learns a consistent scene representation in its encoded latent space, and that the decoder can then be applied to this latent representation in order to synthesize the scene from a novel viewpoint. We evaluate Pix2Shape with experiments on the ShapeNet dataset as well as on a novel benchmark we developed -called 3D-IQTTto evaluate models based on their ability to enable 3d spatial reasoning. Qualitative and quantitative evaluation demonstrate Pix2Shape's ability to solve scene reconstruction, generation and understanding tasks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.