Source author record

Sangxia Huang

Sangxia Huang appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Computational Complexity Computer Vision Discrete Mathematics math.OC

Catalog footprint

What is connected

5works

4topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Bootstrapped Representation Learning for Skeleton-Based Action Recognition

In this work, we study self-supervised representation learning for 3D skeleton-based action recognition. We extend Bootstrap Your Own Latent (BYOL) for representation learning on skeleton sequence data and propose a new data augmentation strategy including two asymmetric transformation pipelines. We also introduce a multi-viewpoint sampling method that leverages multiple viewing angles of the same action captured by different cameras. In the semi-supervised setting, we show that the performance can be further improved by knowledge distillation from wider networks, leveraging once more the unlabeled samples. We conduct extensive experiments on the NTU-60 and NTU-120 datasets to demonstrate the performance of our proposed method. Our method consistently outperforms the current state of the art on both linear evaluation and semi-supervised benchmarks.

preprint2016arXiv

Small Extended Formulation for Knapsack Cover Inequalities from Monotone Circuits

Initially developed for the min-knapsack problem, the knapsack cover inequalities are used in the current best relaxations for numerous combinatorial optimization problems of covering type. In spite of their widespread use, these inequalities yield linear programming (LP) relaxations of exponential size, over which it is not known how to optimize exactly in polynomial time. In this paper we address this issue and obtain LP relaxations of quasi-polynomial size that are at least as strong as that given by the knapsack cover inequalities. For the min-knapsack cover problem, our main result can be stated formally as follows: for any $\varepsilon >0$, there is a $(1/\varepsilon)^{O(1)}n^{O(\log n)}$-size LP relaxation with an integrality gap of at most $2+\varepsilon$, where $n$ is the number of items. Prior to this work, there was no known relaxation of subexponential size with a constant upper bound on the integrality gap. Our construction is inspired by a connection between extended formulations and monotone circuit complexity via Karchmer-Wigderson games. In particular, our LP is based on $O(\log^2 n)$-depth monotone circuits with fan-in~$2$ for evaluating weighted threshold functions with $n$ inputs, as constructed by Beimel and Weinreb. We believe that a further understanding of this connection may lead to more positive results complementing the numerous lower bounds recently proved for extended formulations.

preprint2015arXiv

$2^{(\log N)^{1/10-o(1)}}$ Hardness for Hypergraph Coloring

We show that it is quasi-NP-hard to color 2-colorable 8-uniform hypergraphs with $2^{(\log N)^{1/10-o(1)}}$ colors, where $N$ is the number of vertices. There has been much focus on hardness of hypergraph coloring recently. Guruswami, Håstad, Harsha, Srinivasan and Varma showed that it is quasi-NP-hard to color 2-colorable 8-uniform hypergraphs with $2^{2^{Ω(\sqrt{\log\log N})}}$ colors. Their result is obtained by composing standard Label Cover with an inner-verifier based on Low Degree Long Code, using Reed-Muller code testing results by Dinur and Guruswami. Using a different approach, Khot and Saket constructed a new variant of Label Cover, and composed it with Quadratic Code to show quasi-NP-hardness of coloring 2-colorable 12-uniform hypergraphs with $2^{(\log N)^c}$ colors, for some $c$ around 1/20. Their construction of Label Cover is based on a new notion of superposition complexity for CSP instances. The composition with inner-verifier was subsequently improved by Varma, giving the same hardness result for 8-uniform hypergraphs. Our construction uses both Quadratic Code and Low Degree Long Code, and builds upon the work by Khot and Saket. We present a different approach to construct CSP instances with superposition hardness by observing that when the number of assignments is odd, satisfying a constraint in superposition is the same as "odd-covering" the constraint. We employ Low Degree Long Code in order to keep the construction efficient. In the analysis, we also adapt and generalize one of the key theorems by Dinur and Guruswami in the context of analyzing probabilistically checkable proof systems.

preprint2013arXiv

Improved Hardness of Approximating Chromatic Number

We prove that for sufficiently large K, it is NP-hard to color K-colorable graphs with less than 2^{K^{1/3}} colors. This improves the previous result of K versus K^{O(log K)} in Khot [14].

preprint2010arXiv

From Holant To #CSP And Back: Dichotomy For Holant$^c$ Problems

We explore the intricate interdependent relationship among counting problems, considered from three frameworks for such problems: Holant Problems, counting CSP and weighted H-colorings. We consider these problems for general complex valued functions that take boolean inputs. We show that results from one framework can be used to derive results in another, and this happens in both directions. Holographic reductions discover an underlying unity, which is only revealed when these counting problems are investigated in the complex domain $\mathbb{C}$. We prove three complexity dichotomy theorems, leading to a general theorem for Holant$^c$ problems. This is the natural class of Holant problems where one can assign constants 0 or 1. More specifically, given any signature grid on $G=(V,E)$ over a set ${\mathscr F}$ of symmetric functions, we completely classify the complexity to be in P or #P-hard, according to ${\mathscr F}$, of \[\sum_{σ: E \rightarrow \{0,1\}} \prod_{v\in V} f_v(σ\mid_{E(v)}),\] where $f_v \in {\mathscr F} \cup \{{\bf 0}, {\bf 1}\}$ ({\bf 0}, {\bf 1} are the unary constant 0, 1 functions). Not only is holographic reduction the main tool, but also the final dichotomy can be only naturally stated in the language of holographic transformations. The proof goes through another dichotomy theorem on boolean complex weighted #CSP.