14 June 2023 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Visual scenes are composed of visual concepts and have the property of combinatorial explosion. An important reason for humans to efficiently learn from diverse visual scenes is the ability of compositional perception, and it is desirable for artificial intelligence to have similar abilities. Compositional scene representation learning is a task that enables such abilities. In recent years, various methods have been proposed to apply deep neural networks, which have been proven to be advantageous in representation learning, to learn compositional scene representations via reconstruction, advancing this research direction into the deep learning era. Learning via reconstruction is advantageous because it may utilize massive unlabeled data and avoid costly and laborious data annotation. In this survey, we first outline the current progress on reconstruction-based compositional scene representation learning with deep neural networks, including development history and categorizations of existing methods from the perspectives of the modeling of visual scenes and the inference of scene representations; then provide benchmarks, including an open source toolbox to reproduce the benchmark experiments, of representative methods that consider the most extensively studied problem setting and form the foundation for other methods; and finally discuss the limitations of existing methods and future directions of this research topic.
1 December 2017 in IEEE Communications Letters
Graphics processing units (GPUs) are well-suited for decoding low-density parity-check (LDPC) codes because of their massive parallelisms and high flexibility. Implementations of high-throughput GPU-based LDPC decoders are described in this letter. By applying a novel message updating scheme and reducing shared memory consumption, the flooding-based and layered-based decoders outperform the corresponding state-of-the-art designs, with 55% and 34% improvements of normalized throughputs. On a single GeForce GTX 1080 Ti GPU, the two decoders achieve 4.77 and 3.67 Gb/s throughputs by incorporating early termination (ET) criterion. Throughputs of the GPU-based decoders are comparable with an implementation on the largest density field-programmable gate array (FPGA).
3 January 2024 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Visual scenes are extremely diverse, not only because there are infinite possible combinations of objects and backgrounds but also because the observations of the same scene may vary greatly with the change of viewpoints. When observing a multi-object visual scene from multiple viewpoints, humans can perceive the scene compositionally from each viewpoint while achieving the so-called “object constancy” across different viewpoints, even though the exact viewpoints are untold. This ability is essential for humans to identify the same object while moving and to learn from vision efficiently. It is intriguing to design models that have a similar ability. In this paper, we consider a novel problem of learning compositional scene representations from multiple unspecified (i.e., unknown and unrelated) viewpoints without using any supervision and propose a deep generative model which separates latent representations into a viewpoint-independent part and a viewpoint-dependent part to solve this problem. During the inference, latent representations are randomly initialized and iteratively updated by integrating the information in different viewpoints with neural networks. Experiments on several specifically designed synthetic datasets have shown that the proposed method can effectively learn from multiple unspecified viewpoints.