georgecarlyle76

9 ms·

31.
How we update Alexa's skill selector without causing “catastrophic forgetting”(developer.amazon.com)
6 points by georgecarlyle76 8y ago|0 comments
32.
“Alexa, turn down the lights and play music”: How to handle compound requests(developer.amazon.com)
8 points by georgecarlyle76 8y ago|0 comments
33.
Multi-speaker TTS models outperform single-speaker ones but sound the same(developer.amazon.com)
5 points by georgecarlyle76 8y ago|0 comments
34.
Using wake word acoustics to filter background speech improves speech recognizer(developer.amazon.com)
7 points by georgecarlyle76 8y ago|0 comments
35.
Alexa uses tri-training, “pyramidal” NNs for sound recognition, media detection(developer.amazon.com)
5 points by georgecarlyle76 8y ago|0 comments
36.
Cross-Lingual Transfer Learning Halves Data Requirements to Bootstrap AI Systems(developer.amazon.com)
7 points by georgecarlyle76 8y ago|0 comments
37.
Alexa researchers train speech recognizer on 1M hours of unlabeled speech(developer.amazon.com)
12 points by georgecarlyle76 8y ago|1 comments
38.
Joint training on beamforming/ASR improves Alexa's ASR, with fewer microphones(developer.amazon.com)
3 points by georgecarlyle76 8y ago|0 comments
39.
Audio Watermarking Algorithm First to Solve Second-Screen Problem in Real Time(developer.amazon.com)
10 points by georgecarlyle76 8y ago|0 comments
40.
Adversarial Training Produces Synthetic Data for Machine Learning(developer.amazon.com)
10 points by georgecarlyle76 8y ago|0 comments
41.
Machine-Labeled Data plus Artificial Noise = Better Speech Recognition(developer.amazon.com)
6 points by georgecarlyle76 8y ago|0 comments
42.
To Correct Imbalances in Training Data, Don’t Oversample: Cluster(developer.amazon.com)
5 points by georgecarlyle76 8y ago|0 comments
43.
Conversational-AI Innovations from Contestants in the 2018 Alexa Prize(developer.amazon.com)
7 points by georgecarlyle76 8y ago|0 comments
44.
The AI Tools That Amazon Built to Let Alexa Prize Participants Focus on Science(developer.amazon.com)
1 points by georgecarlyle76 8y ago|0 comments
45.
Why Alexa Won't Wake Up When She Hears Her Name in Amazon's Super Bowl Ad(developer.amazon.com)
6 points by georgecarlyle76 8y ago|0 comments
46.
Transfer learning trains NN to recognize new classes, w/o original training data(developer.amazon.com)
6 points by georgecarlyle76 8y ago|0 comments
47.
With new kernel method, NLP system learns in four hours what used to take 7 days(developer.amazon.com)
17 points by georgecarlyle76 8y ago|0 comments
48.
Leveraging Unannotated Data to Bootstrap Alexa Functions More Quickly(developer.amazon.com)
5 points by georgecarlyle76 8y ago|0 comments
49.
New Method for Compressing Neural Networks Better Preserves Accuracy(developer.amazon.com)
120 points by georgecarlyle76 8y ago|36 comments
50.
How to use the (non-differentiable) F1 score to train a machine learning model(developer.amazon.com)
4 points by georgecarlyle76 8y ago|0 comments
51.
How Alexa knows “peanut butter” is one shopping-list item, not two(developer.amazon.com)
246 points by georgecarlyle76 8y ago|140 comments
52.
New data representation scheme reduces Alexa's skill selection errors by 40%(developer.amazon.com)
8 points by georgecarlyle76 8y ago|0 comments
53.
Building language models directly from grammars reduces ASR error up to 15%(developer.amazon.com)
3 points by georgecarlyle76 8y ago|0 comments
54.
How NLU domains can independently calibrate their decisions to field a query(developer.amazon.com)
3 points by georgecarlyle76 8y ago|0 comments
55.
Amazon scientists: What NeurIPS (formerly NIPS) tells us about the state of AI(blog.aboutamazon.com)
1 points by georgecarlyle76 8y ago|0 comments
56.
Varying Speaking Styles with Neural Text-to-Speech(developer.amazon.com)
86 points by georgecarlyle76 8y ago|12 comments
57.
New details about how Alexa decides which skill will handle a request(developer.amazon.com)
1 points by georgecarlyle76 8y ago|0 comments
58.
New details about the model that decides which Alexa skill will handle a request(developer.amazon.com)
7 points by georgecarlyle76 8y ago|0 comments
59.
Researchers at Amazon and U. of Sheffield team on project to detect fake news(developer.amazon.com)
4 points by georgecarlyle76 8y ago|0 comments
60.
Locating sound with an Echo's microphone array could enable clap-based controls(developer.amazon.com)
4 points by georgecarlyle76 8y ago|0 comments

More ›