LSTMVis - Visual Analysis for Recurrent Neural Networks
LSTMVis a visual analysis tool for recurrent neural networks with a focus on understanding these hidden state dynamics. The tool allows a user to select a hypothesis input range to focus on local state changes, to match these states changes to similar patterns in a large data set, and to align these results with structural annotations from their domain. We provide data for the tool to analyze specific hidden state properties on dataset containing nesting, phrase structure, and chord progressions, and demonstrate how the tool can be used to isolate patterns for further statistical analysis.
http://lstm.seas.harvard.edu/
LSTMVis a visual analysis tool for recurrent neural networks with a focus on understanding these hidden state dynamics. The tool allows a user to select a hypothesis input range to focus on local state changes, to match these states changes to similar patterns in a large data set, and to align these results with structural annotations from their domain. We provide data for the tool to analyze specific hidden state properties on dataset containing nesting, phrase structure, and chord progressions, and demonstrate how the tool can be used to isolate patterns for further statistical analysis.
http://lstm.seas.harvard.edu/
Temporal Activity Detection in Untrimmed Videos with Recurrent Neural Networks
This project is a baseline in the activity classification and its temporal location, focused on the ActivityNet Challenge. Here is detailed all the process of our proposed pipeline, as well the trained models and the utility to classify and temporally localize activities on new videos given. All the steps have been detailed, from downloading the dataset, to predicting the temporal locations going through the feature extraction and also the training.
https://imatge-upc.github.io/activitynet-2016-cvprw/
https://github.com/imatge-upc/activitynet-2016-cvprw
https://www.youtube.com/watch?v=3G-Vdmsluw0
#keras #video #activity #cvprw
This project is a baseline in the activity classification and its temporal location, focused on the ActivityNet Challenge. Here is detailed all the process of our proposed pipeline, as well the trained models and the utility to classify and temporally localize activities on new videos given. All the steps have been detailed, from downloading the dataset, to predicting the temporal locations going through the feature extraction and also the training.
https://imatge-upc.github.io/activitynet-2016-cvprw/
https://github.com/imatge-upc/activitynet-2016-cvprw
https://www.youtube.com/watch?v=3G-Vdmsluw0
#keras #video #activity #cvprw
FACE DETECTION BY LITERATURE
Обзор и бенчмарки различных методов детектирования лиц
#facedetection #cv
http://www.erogol.com/face-detection-networks-literature/
Обзор и бенчмарки различных методов детектирования лиц
#facedetection #cv
http://www.erogol.com/face-detection-networks-literature/
Блог Kaggle: Что мы читаем, 15 Любимых Data Science ресурсов
http://blog.kaggle.com/2016/09/13/what-were-reading-data-science-resources/
http://blog.kaggle.com/2016/09/13/what-were-reading-data-science-resources/
Google, Facebook, Amazon объединяют силы для развития искусственного разума
http://www.bbc.com/russian/news-37503129
http://www.bbc.com/russian/news-37503129
YouTube-8M: датасет, это как ImagNet только для видео!
4800 классов и 8 миллионов видео
YouTube-8M is a large-scale labeled video dataset that consists of 8 million YouTube video IDs and associated labels from a diverse vocabulary of 4800 visual entities. It also comes with precomputed state-of-the-art vision features from billions of frames, which fit on a single hard disk. This makes it possible to train video models from hundreds of thousands of video hours in less than a day on 1 GPU!
https://research.google.com/youtube8m/
http://arxiv.org/pdf/1609.08675v1.pdf
#dataset
4800 классов и 8 миллионов видео
YouTube-8M is a large-scale labeled video dataset that consists of 8 million YouTube video IDs and associated labels from a diverse vocabulary of 4800 visual entities. It also comes with precomputed state-of-the-art vision features from billions of frames, which fit on a single hard disk. This makes it possible to train video models from hundreds of thousands of video hours in less than a day on 1 GPU!
https://research.google.com/youtube8m/
http://arxiv.org/pdf/1609.08675v1.pdf
#dataset
Large Scale Movie Denoscription and Understanding Challenge (LSMDC), at ECCV 2016
The challenge will be presented at the "Joint 2nd Workshop on Storytelling with Images and Videos (VisStory) and Large Scale Movie Denoscription and Understanding Challenge (LSMDC 2016)" in conjunction with ECCV 2016, Amsterdam, The Netherlands.
The challenge will be presented at the "Joint 2nd Workshop on Storytelling with Images and Videos (VisStory) and Large Scale Movie Denoscription and Understanding Challenge (LSMDC 2016)" in conjunction with ECCV 2016, Amsterdam, The Netherlands.