Zobrazeno 1 - 10
of 107
pro vyhledávání: '"Kender, John"'
We demonstrate the efficiencies and explanatory abilities of extensions to the common tools of Autoencoders and LLM interpreters, in the novel context of comparing different cultural approaches to the same international news event. We develop a new C
Externí odkaz:
http://arxiv.org/abs/2408.07791
Transfer learning is a deep-learning technique that ameliorates the problem of learning when human-annotated labels are expensive and limited. In place of such labels, it uses instead the previously trained weights from a well-chosen source model as
Externí odkaz:
http://arxiv.org/abs/2207.03554
Human motion prediction and understanding is a challenging problem. Due to the complex dynamic of human motion and the non-deterministic aspect of future prediction. We propose a novel sequence-to-sequence model for human motion prediction and featur
Externí odkaz:
http://arxiv.org/abs/2012.15378
Autor:
Bhattacharjee, Bishwaranjan, Kender, John R., Hill, Matthew, Dube, Parijat, Huo, Siyu, Glass, Michael R., Belgodere, Brian, Pankanti, Sharath, Codella, Noel, Watson, Patrick
Transfer learning enhances learning across tasks, by leveraging previously learned representations -- if they are properly chosen. We describe an efficient method to accurately estimate the appropriateness of a previously trained model for use in a n
Externí odkaz:
http://arxiv.org/abs/1908.07630
Autor:
Xu, Qiangeng, Kender, John
In the research of the impact of gestures using by a lecturer, one challenging task is to infer the attention of a group of audiences. Two important measurements that can help infer the level of attention are eye movement data and Electroencephalogra
Externí odkaz:
http://arxiv.org/abs/1712.09709
Predicting and understanding human motion dynamics has many applications, such as motion synthesis, augmented reality, security, and autonomous vehicles. Due to the recent success of generative adversarial networks (GAN), there has been much interest
Externí odkaz:
http://arxiv.org/abs/1711.09561
Social information networks, such as YouTube, contains traces of both explicit online interaction (such as "like", leaving a comment, or subscribing to video feed), and latent interactions (such as quoting, or remixing parts of a video). We propose v
Externí odkaz:
http://arxiv.org/abs/1210.0623
Autor:
Haubold, Alexander, Kender, John R.
We introduce a novel and inexpensive approach for the temporal alignment of speech to highly imperfect transcripts from automatic speech recognition (ASR). Transcripts are generated for extended lecture and presentation videos, which in some cases fe
Externí odkaz:
http://arxiv.org/abs/cs/0612139
Autor:
Haubold, Alexander, Kender, John R.
We investigate the symmetric Kullback-Leibler (KL2) distance in speaker clustering and its unreported effects for differently-sized feature matrices. Speaker data is represented as Mel Frequency Cepstral Coefficient (MFCC) vectors, and features are c
Externí odkaz:
http://arxiv.org/abs/cs/0612138
Autor:
Haubold, Alexander, Kender, John R.
We investigate methods of segmenting, visualizing, and indexing presentation videos by separately considering audio and visual data. The audio track is segmented by speaker, and augmented with key phrases which are extracted using an Automatic Speech
Externí odkaz:
http://arxiv.org/abs/cs/0501044