arXiv subjects

Jack Wang

Publications and source records attributed to Jack Wang.

9 recordsLinked to original sources

arXiv2026

A Glance Is All You Need: Single-Pass Fine-Grained Image Captioning with SimLoss

An image may be worth a thousand words, but most captioning models describe it in only a few. Modern vision-language models produce fluent high-level captions, yet routinely miss the attributes, counts, textures, materials, and spatial relations that make an image visually specific. Recent multi-stage systems recover some of these details through generation, decomposition, verification, and rewriting, but they do so at the expense of substantially higher inference latency. We propose SimLoss, a reference-free embedding-space objective for single-pass fine-grained image captioning. SimLoss trains a vision-language model to align its projected hidden-state representation with a frozen image embedding through an InfoNCE contrastive loss, supplying a dense visual supervision signal before any text is decoded, and requiring neither human-written fine-grained captions nor pseudo-captions from a multi-stage pipeline. We instantiate it as SimLoss FFT, which backpropagates through a locally available embedding model, and SimLoss GRPO, which treats that model as a black-box reward. Compared with single-pass, multi-stage verification, reward-optimized, and perception-aware baselines, the fully differentiable fine-tuning variant, SimLoss FFT, achieves the highest precision while nearly matching the F1 score of the multi-stage method, all while retaining single-pass inference and running roughly 20 times faster than the multi-stage pipeline. The reward-based variant SimLoss GRPO attains the strongest recall. Together, these results show that embedding-space supervision can recover the quality of multi-stage verification at the latency of a single-pass captioner.

cs.CV↗

arXiv2024

Tangent Space of the Stable And Unstable Manifold of Anosov Diffeomorphism on 2-Torus

In this paper we describe the tangent vectors of the stable and unstable manifold of a class of Anosov diffeomorphisms on the torus $\mathbb{T}^2$ using the method of formal series and derivative trees. We start with linear automorphism that is hyperbolic and whose eigenvectors are orthogonal. Then we study the perturbation of such maps by trigonometric polynomial. It is known that there exist a (continuous) map $H$ which acts as a change of coordinate between the perturbed and unperturbed system, but such a map is in general, not differentiable. By "re-scaling" the parametrization $H$, we will be able to obtain the explicit formula for the tangent vectors of these maps.

math.DS↗

arXiv2019

Calibration of free-space and fiber-coupled single-photon detectors

We measure the detection efficiency of single-photon detectors at wavelengths near 851 nm and 1533.6 nm. We investigate the spatial uniformity of one free-space-coupled single-photon avalanche diode and present a comparison between fusion-spliced and connectorized fiber-coupled single-photon detectors. We find that our expanded relative uncertainty for a single measurement of the detection efficiency is as low as 0.70 % for fiber-coupled measurements at 1533.6 nm and as high as 1.78 % for our free-space characterization at 851.7 nm. The detection-efficiency determination includes corrections for afterpulsing, dark count, and count-rate effects of the single-photon detector with the detection efficiency interpolated to operation at a specified detected count rate.

quant-ph↗

arXiv2018

The Prophet Inequality Can Be Solved Optimally with a Single Set of Samples

The setting of the classic prophet inequality is as follows: a gambler is shown the probability distributions of $n$ independent, non-negative random variables with finite expectations. In their indexed order, a value is drawn from each distribution, and after every draw the gambler may choose to accept the value and end the game, or discard the value permanently and continue the game. What is the best performance that the gambler can achieve in comparison to a prophet who can always choose the highest value? Krengel, Sucheston, and Garling solved this problem in 1978, showing that there exists a strategy for which the gambler can achieve half as much reward as the prophet in expectation. Furthermore, this result is tight. In this work, we consider a setting in which the gambler is allowed much less information. Suppose that the gambler can only take one sample from each of the distributions before playing the game, instead of knowing the full distributions. We provide a simple and intuitive algorithm that recovers the original approximation of $\frac{1}{2}$. Our algorithm works against even an almighty adversary who always chooses a worst-case ordering, rather than the standard offline adversary. The result also has implications for mechanism design -- there is much interest in designing competitive auctions with a finite number of samples from value distributions rather than full distributional knowledge.

cs.DS↗

arXiv2018

A PRQ Search Method for Probabilistic Objects

This article proposes an PQR search method for probabilistic objects. The main idea of our method is to use a strategy called \textit{pre-approximation} that can reduce the initial problem to a highly simplified version, implying that it makes the rest of steps easy to tackle. In particular, this strategy itself is pretty simple and easy to implement. Furthermore, motivated by the cost analysis, we further optimize our solution. The optimizations are mainly based on two insights: (\romannumeral 1) the number of \textit{effective subdivision}s is no more than 1; and (\romannumeral 2) an entity with the larger \textit{span} is more likely to subdivide a single region. We demonstrate the effectiveness and efficiency of our proposed approaches through extensive experiments under various experimental settings.

cs.DB↗

arXiv2018

Finding Efficient Region in The Plane with Line segments

Let $\mathscr O$ be a set of $n$ disjoint obstacles in $\mathbb{R}^2$, $\mathscr M$ be a moving object. Let $s$ and $l$ denote the starting point and maximum path length of the moving object $\mathscr M$, respectively. Given a point $p$ in ${R}^2$, we say the point $p$ is achievable for $\mathscr M$ such that $π(s,p)\leq l$, where $π(\cdot)$ denotes the shortest path length in the presence of obstacles. One is to find a region $\mathscr R$ such that, for any point $p\in \mathbb{R}^2$, if it is achievable for $\mathscr M$, then $p\in \mathscr R$; otherwise, $p\notin \mathscr R$. In this paper, we restrict our attention to the case of line-segment obstacles. To tackle this problem, we develop three algorithms. We first present a simpler-version algorithm for the sake of intuition. Its basic idea is to reduce our problem to computing the union of a set of circular visibility regions (CVRs). This algorithm takes $O(n^3)$ time. By analysing its dominant steps, we break through its bottleneck by using the short path map (SPM) technique to obtain those circles (unavailable beforehand), yielding an $O(n^2\log n)$ algorithm. Owing to the finding above, the third algorithm also uses the SPM technique. It however, does not continue to construct the CVRs. Instead, it directly traverses each region of the SPM to trace the boundaries, the final algorithm obtains $O(n\log n)$ complexity.

cs.DS↗

arXiv2018

A Simple Algorithm for Computing BOCP

In this article, we devise a concise algorithm for computing BOCP. Our method is simple, easy-to-implement but without loss of efficiency. Given two circular-arc polygons with $m$ and $n$ edges respectively, our method runs in $O(m+n+(l+k)\log l)$ time, using $O(m+n+k)$ space, where $k$ is the number of intersections, and $l$ is the number of {edge}s. Our algorithm has the power to approximate to linear complexity when $k$ and $l$ are small. The superiority of the proposed algorithm is also validated through empirical study.

cs.DS↗

arXiv2018

An Improved Solution for Restricted and Uncertain TRQ

CSPTRQ is an interesting problem and its has attracted much attention. The CSPTRQ is a variant of the traditional PTRQ. As objects moving in a constrained-space are common, clearly, it can also find many applications. At the first sight, our problem can be easily tackled by extending existing methods used to answer the PTRQ. Unfortunately, those classical techniques are not well suitable for our problem, due to a set of new challenges. We develop targeted solutions and demonstrate the efficiency and effectiveness of the proposed methods through extensive experiments.

cs.DB↗

arXiv2018

DeepHeart: Semi-Supervised Sequence Learning for Cardiovascular Risk Prediction

We train and validate a semi-supervised, multi-task LSTM on 57,675 person-weeks of data from off-the-shelf wearable heart rate sensors, showing high accuracy at detecting multiple medical conditions, including diabetes (0.8451), high cholesterol (0.7441), high blood pressure (0.8086), and sleep apnea (0.8298). We compare two semi-supervised train- ing methods, semi-supervised sequence learning and heuristic pretraining, and show they outperform hand-engineered biomarkers from the medical literature. We believe our work suggests a new approach to patient risk stratification based on cardiovascular risk scores derived from popular wearables such as Fitbit, Apple Watch, or Android Wear.

cs.LG↗