dynamic time warping
PulseAugur coverage of dynamic time warping — every cluster mentioning dynamic time warping across labs, papers, and developer communities, ranked by signal.
3 天有情绪数据
-
New probabilistic framework enhances data alignment with uncertainty modeling
Researchers have developed a new probabilistic framework called uncertainty-DTW (uDTW) for aligning structured data, enhancing traditional methods like Dynamic Time Warping. This approach models pairwise correspondences…
-
深度神经网络框架评估眼动反应时间用于轻度TBI评估
研究人员开发了一个新颖的框架，将脑电图（EEG）与增强现实（AR）前庭/眼动筛查（VOMS）任务相结合，以估计眼动反应时间。该系统利用冗余离散小波变换（RDWT）驱动的深度神经网络来分析EEG信号，这是一种有效的去噪策略。然后采用动态时间规整（DTW）来计算反应时间，揭示了显著的受试间差异和任务依赖的时间行为，表明其在多模态轻度创伤性脑损伤（mTBI）评估中的潜力。
-
New method uses path signatures for efficient online goal recognition
Researchers have developed a new method for online goal recognition that utilizes path signatures from rough path theory. This approach efficiently encodes and compares large trajectory datasets, outperforming existing …
-
New defense offers certified robustness for time-series anomaly detection
Researchers have developed the first defense mechanism that provides certified robustness for time-series anomaly detection under the Dynamic Time Warping (DTW) metric. This new approach adapts the randomized smoothing …
-
Researchers improve medical VQA with trajectory-aware process supervision
Researchers have developed a novel method to improve medical visual question answering (VQA) systems by incorporating trajectory-aware process supervision. This approach utilizes a two-stage training framework, starting…
-
TokenTiming: A Dynamic Alignment Method for Universal Speculative Decoding Model Pairs
Researchers have developed a new method called TokenTiming, inspired by Dynamic Time Warping, to improve the efficiency of speculative decoding in large language models. This technique allows for the use of draft and ta…
-
New Soft-MSM method offers improved time series alignment and clustering
Researchers have developed Soft-MSM, a new differentiable loss function for time series analysis that improves upon existing methods like Soft-DTW. Soft-MSM incorporates context-aware transition costs, making it more ef…
-
ESPADA框架将机器人模仿学习速度提升2倍
研究人员开发了ESPADA，一个旨在通过智能下采样演示数据来加速机器人操作任务的新框架。ESPADA利用VLM-LLM管道来识别和保留机器人动作的关键阶段，同时积极加速非关键部分。这种方法在无需重新训练或额外数据的情况下，实现了约两倍的执行速度提升，并在模拟和真实世界实验中保持了高成功率。