Author name cluster

Jian Ding

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

AAAI Conference 2024 Conference Paper

Prompting Segmentation with Sound Is Generalizable Audio-Visual Source Localizer

Yaoting Wang
Weisong Liu
Guangyao Li
Jian Ding
Di Hu
Xi Li

Never having seen an object and heard its sound simultaneously, can the model still accurately localize its visual position from the input audio? In this work, we concentrate on the Audio-Visual Localization and Segmentation tasks but under the demanding zero-shot and few-shot scenarios. To achieve this goal, different from existing approaches that mostly employ the encoder-fusion-decoder paradigm to decode localization information from the fused audio-visual feature, we introduce the encoder-prompt-decoder paradigm, aiming to better fit the data scarcity and varying data distribution dilemmas with the help of abundant knowledge from pre-trained models. Specifically, we first propose to construct a Semantic-aware Audio Prompt (SAP) to help the visual foundation model focus on sounding objects, meanwhile, the semantic gap between the visual and audio modalities is also encouraged to shrink. Then, we develop a Correlation Adapter (ColA) to keep minimal training efforts as well as maintain adequate knowledge of the visual foundation model. By equipping with these means, extensive experiments demonstrate that this new paradigm outperforms other fusion-based methods in both the unseen class and cross-dataset settings. We hope that our work can further promote the generalization study of Audio-Visual Localization and Segmentation in practical application scenarios. Project page: https://github.com/GeWu-Lab/Generalizable-Audio-Visual-Segmentation

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

VRSBench: A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding

Xiang Li
Jian Ding
Mohamed Elhoseiny

We introduce a new benchmark designed to advance the development of general-purpose, large-scale vision-language models for remote sensing images. Although several vision-language datasets in remote sensing have been proposed to pursue this goal, existing datasets are typically tailored to single tasks, lack detailed object information, or suffer from inadequate quality control. Exploring these improvement opportunities, we present a Versatile vision-language Benchmark for Remote Sensing image understanding, termed VRSBench. This benchmark comprises 29, 614 images, with 29, 614 human-verified detailed captions, 52, 472 object references, and 123, 221 question-answer pairs. It facilitates the training and evaluation of vision-language models across a broad spectrum of remote sensing image understanding tasks. We further evaluated state-of-the-art models on this benchmark for three vision-language tasks: image captioning, visual grounding, and visual question answering. Our work aims to significantly contribute to the development of advanced vision-language models in the field of remote sensing. The data and code can be accessed at https: //vrsbench. github. io.

PDF Details DOI

AAAI Conference 2023 Conference Paper

Few-Shot Object Detection via Variational Feature Aggregation

Jiaming Han
Yuqiang Ren
Jian Ding
Ke Yan
Gui-Song Xia

As few-shot object detectors are often trained with abundant base samples and fine-tuned on few-shot novel examples, the learned models are usually biased to base classes and sensitive to the variance of novel examples. To address this issue, we propose a meta-learning framework with two novel feature aggregation schemes. More precisely, we first present a Class-Agnostic Aggregation (CAA) method, where the query and support features can be aggregated regardless of their categories. The interactions between different classes encourage class-agnostic representations and reduce confusion between base and novel classes. Based on the CAA, we then propose a Variational Feature Aggregation (VFA) method, which encodes support examples into class-level support features for robust feature aggregation. We use a variational autoencoder to estimate class distributions and sample variational features from distributions that are more robust to the variance of support examples. Besides, we decouple classification and regression tasks so that VFA is performed on the classification branch without affecting object localization. Extensive experiments on PASCAL VOC and COCO demonstrate that our method significantly outperforms a strong baseline (up to 16%) and previous state-of-the-art methods (4% in average).

PDF Details DOI

STOC Conference 2019 Conference Paper

Capacity lower bound for the Ising perceptron

Jian Ding
Nike Sun

We consider the Ising perceptron with gaussian disorder, which is equivalent to the discrete cube {−1,+1} N intersected by M random half-spaces. The perceptron’s capacity is the largest integer M N for which the intersection is nonempty. It is conjectured by Krauth and Mézard (1989) that the (random) ratio M N / N converges in probability to an explicit constant α ⋆ ≐ 0.83. Kim and Roche (1998) proved the existence of a positive constant γ such that γ ≤ M N / N ≤ 1−γ with high probability; see also Talagrand (1999). In this paper we show that the Krauth–Mézard conjecture α ⋆ is a lower bound with positive probability, under the condition that an explicit univariate function S (λ) is maximized at λ=0. Our proof is an application of the second moment method to a certain slice of perceptron configurations, as selected by the so-called TAP (Thouless, Anderson, and Palmer, 1977) or AMP (approximate message passing) iteration, whose scaling limit has been characterized by Bayati and Montanari (2011) and Bolthausen (2012). For verifying the condition on S (λ) we outline one approach, which is implemented in the current version using (nonrigorous) numerical integration packages. In a future version of this paper we intend to complete the verification by implementing a rigorous numerical method.

Details

STOC Conference 2015 Conference Paper

Proof of the Satisfiability Conjecture for Large k

Jian Ding
Allan Sly
Nike Sun

We establish the satisfiability threshold for random k-SAT for all k ≥ k 0 . That is, there exists a limiting density α s (k) such that a random k-SAT formula of clause density α is with high probability satisfiable for α α s . The satisfiability threshold α s is given explicitly by the one-step replica symmetry breaking (1SRB) prediction from statistical physics. We believe that our methods may apply to a range of random constraint satisfaction problems in the 1RSB class.

Details

STOC Conference 2014 Conference Paper

Bandits with switching costs: T 2/3 regret

Ofer Dekel
Jian Ding
Tomer Koren
Yuval Peres

We study the adversarial multi-armed bandit problem in a setting where the player incurs a unit cost each time he switches actions. We prove that the player's T -round minimax regret in this setting is [EQUATION], thereby closing a fundamental gap in our understanding of learning with bandit feedback. In the corresponding full-information version of the problem, the minimax regret is known to grow at a much slower rate of Θ(√ T ). The difference between these two rates provides the first indication that learning with bandit feedback can be significantly harder than learning with full information feedback (previous results only showed a different dependence on the number of actions, but not on T .) In addition to characterizing the inherent difficulty of the multi-armed bandit problem with switching costs, our results also resolve several other open problems in online learning. One direct implication is that learning with bandit feedback against bounded-memory adaptive adversaries has a minimax regret of [EQUATION]. Another implication is that the minimax regret of online learning in adversarial Markov decision processes (MDPs) is [EQUATION]. The key to all of our results is a new randomized construction of a multi-scale random walk, which is of independent interest and likely to prove useful in additional settings.

Details

STOC Conference 2014 Conference Paper

Satisfiability threshold for random regular NAE-SAT

Jian Ding
Allan Sly
Nike Sun

We consider the random regular k -nae-sat problem with n variables each appearing in exactly d clauses. For all k exceeding an absolute constant k 0 , we establish explicitly the satisfiability threshold d * ∈ d * ( k ). We prove that for d d * the problem is unsatisfiable with high probability. If the threshold d * lands exactly on an integer, we show that the problem is satisfiable with probability bounded away from both zero and one. This is the first result to locate the exact satisfiability threshold in a random constraint satisfaction problem exhibiting the condensation phenomenon identified by Krzakał a et al. (2007). Our proof verifies the onestep replica symmetry breaking formalism for this model. We expect our methods to be applicable to a broad range of random constraint satisfaction problems and combinatorial problems on random graphs.

Details

STOC Conference 2011 Conference Paper

Cover times, blanket times, and majorizing measures

Jian Ding
James R. Lee
Yuval Peres

We exhibit a strong connection between cover times of graphs, Gaussian processes, and Talagrand's theory of majorizing measures. In particular, we show that the cover time of any graph G is equivalent, up to universal constants, to the square of the expected maximum of the Gaussian free field on G, scaled by the number of edges in G. This allows us to resolve a number of open questions. We give a deterministic polynomial-time algorithm that computes the cover time to within an O(1) factor for any graph, answering a question of Aldous and Fill (1994). We also positively resolve the blanket time conjectures of Winkler and Zuckerman (1996), showing that for any graph, the blanket and cover times are within an O(1) factor. The best previous approximation factor for both these problems was O((log log n) 2 ) for n-vertex graphs, due to Kahn, Kim, Lovasz, and Vu (2000).

Details