arXiv subjects

Ting-Wei Lin

Publications and source records attributed to Ting-Wei Lin.

6 recordsLinked to original sources

arXiv2026

Event ActivityNet: A Large-Scale Simulated-Event Benchmark for Untrimmed Action Understanding

Long-horizon event-based action understanding remains underexplored because existing datasets largely comprise short, trimmed clips, while collecting native event streams with dense temporal annotations is costly. We introduce Event ActivityNet, a large-scale simulated-event benchmark derived from human-annotated, untrimmed ActivityNet videos. It comprises 3,263 videos, 200 action classes, and 106.94 hours, with matched 5-bin and 9-bin event-voxel representations, temporal action annotations, and timestamped captions. The benchmark supports annotated-segment action recognition, auxiliary event-language alignment, and causal online temporal action localization. We generate event voxels directly from non-interpolated source videos in decoded frame order, retain per-video rational nominal or average frame-rate metadata for approximate time mapping, and use action-center reconstruction LPIPS as a soft diagnostic of retained reconstructable content. We establish baselines for adaptive event framing, prompt-caption alignment, and event-only, RGB-only, and RGB-event localization. Under a progressive nested-scale training protocol, recognition Top-1 accuracy increases from 52.25 to 66.42, while online temporal localization average mAP improves from 21.7 to 29.0. Moreover, staged Event ActivityNet pretraining followed by native-event fine-tuning consistently outperforms target-only and joint-from-scratch training across multiple supervision budgets. Event ActivityNet provides a scalable benchmark for long-horizon event modeling, although native-camera evaluation remains essential for deployment-oriented conclusions.

cs.CV↗

arXiv2024

A Study on Synthesizing Expressive Violin Performances: Approaches and Comparisons

Expressive music synthesis (EMS) for violin performance is a challenging task due to the disagreement among music performers in the interpretation of expressive musical terms (EMTs), scarcity of labeled recordings, and limited generalization ability of the synthesis model. These challenges create trade-offs between model effectiveness, diversity of generated results, and controllability of the synthesis system, making it essential to conduct a comparative study on EMS model design. This paper explores two violin EMS approaches. The end-to-end approach is a modification of a state-of-the-art text-to-speech generator. The parameter-controlled approach is based on a simple parameter sampling process that can render note lengths and other parameters compatible with MIDI-DDSP. We study these two approaches (in total, three model variants) through objective and subjective experiments and discuss several key issues of EMS based on the results.

cs.SD↗

arXiv2024

MOSA: Music Motion with Semantic Annotation Dataset for Cross-Modal Music Processing

In cross-modal music processing, translation between visual, auditory, and semantic content opens up new possibilities as well as challenges. The construction of such a transformative scheme depends upon a benchmark corpus with a comprehensive data infrastructure. In particular, the assembly of a large-scale cross-modal dataset presents major challenges. In this paper, we present the MOSA (Music mOtion with Semantic Annotation) dataset, which contains high quality 3-D motion capture data, aligned audio recordings, and note-by-note semantic annotations of pitch, beat, phrase, dynamic, articulation, and harmony for 742 professional music performances by 23 professional musicians, comprising more than 30 hours and 570 K notes of data. To our knowledge, this is the largest cross-modal music dataset with note-level annotations to date. To demonstrate the usage of the MOSA dataset, we present several innovative cross-modal music information retrieval (MIR) and musical content generation tasks, including the detection of beats, downbeats, phrase, and expressive contents from audio, video and motion data, and the generation of musicians' body motion from given music audio. The dataset and codes are available alongside this publication (https://github.com/yufenhuang/MOSA-Music-mOtion-and-Semantic-Annotation-dataset).

cs.SD↗

arXiv2021

Towards a more efficient approach for the satisfiability of two-variable logic

We revisit the satisfiability problem for two-variable logic, denoted by SAT(FO2), which is known to be NEXP-complete. The upper bound is usually derived from its well known Exponential Size Model (ESM) property. Whether it can be determinized efficiently is still an open question. In this paper we present a different approach by reducing it to a novel graph-theoretic problem that we call Conditional Independent Set (CIS). We show that CIS is NP-complete and present two simple algorithms for it with run time O(1.4423^n) and O(1.6181^n), where n is the number of vertices in the graph. We also show that unless the "Strong Exponential Time Hypothesis" (SETH) fails, there is no algorithm for CIS with run time O(1.4141^n). We show that without the equality predicate SAT(FO2) is in fact equivalent to CIS in succinct representation. This yields two algorithms for SAT(FO2) without the equality predicate with run time O(1.4423^{2^n}) and O(1.6181^{2^n}), where n is the number of predicates. To the best of our knowledge, these are the first exact algorithms for an NEXP-complete decidable logic with run time significantly lower than O(2^{2^n}). We also identify a few lower complexity fragments of FO2 which correspond to the tractable fragments of CIS. Similar to CIS, unless SETH fails, there is no algorithm for SAT(FO2) with run time O(1.4141^{2^n}). For the fragment with the equality predicate, we present a linear time many-one reduction to the fragment without the equality predicate. The reduction yields equi-satisfiable formulas with a small constant blow-up in the number of predicates. Finally, we also perform some small experiments which show that our approach is indeed more promising than the existing method (based on the ESM property). The experiments also show that although theoretically it has the worse run time, the second algorithm in general performs better than the first one.

cs.LO↗

arXiv2020

Personalized TV Recommendation: Fusing User Behavior and Preferences

In this paper, we propose a two-stage ranking approach for recommending linear TV programs. The proposed approach first leverages user viewing patterns regarding time and TV channels to identify potential candidates for recommendation and then further leverages user preferences to rank these candidates given textual information about programs. To evaluate the method, we conduct empirical studies on a real-world TV dataset, the results of which demonstrate the superior performance of our model in terms of both recommendation accuracy and time efficiency.

cs.IR↗

arXiv2017

Distributed Training Large-Scale Deep Architectures

Scale of data and scale of computation infrastructures together enable the current deep learning renaissance. However, training large-scale deep architectures demands both algorithmic improvement and careful system configuration. In this paper, we focus on employing the system approach to speed up large-scale training. Via lessons learned from our routine benchmarking effort, we first identify bottlenecks and overheads that hinter data parallelism. We then devise guidelines that help practitioners to configure an effective system and fine-tune parameters to achieve desired speedup. Specifically, we develop a procedure for setting minibatch size and choosing computation algorithms. We also derive lemmas for determining the quantity of key components such as the number of GPUs and parameter servers. Experiments and examples show that these guidelines help effectively speed up large-scale deep learning training.

cs.DC↗