Video Activity Recognition and NLP Q&A Model Example
1 like353 views
The document discusses various applications of deep learning including video, medical imaging, question answering, and summarizes some of the challenges such as ambiguity, sarcasm. It provides an overview of recurrent neural networks for processing sequences of variable length and their use for question answering tasks. Examples are given using a bAbI question answering task to locate people in stories and sequences of words are modeled with parameter sharing.
Video Activity Recognition and NLP Q&A Model Example
- 1. Video
- 2. Applications
- 4. Security
- 6. Challenges
- 7. Storage
- 8. Bandwidth
- 9. Processing
- 10. neon
- 11. UCF-101
- 16. Input Hidden Hidden Output
- 18. Fully Connected
- 20. 3D Volumes
- 23. Demo
- 28. Applications
- 31. Challenges
- 32. Ambiguity John went to the bank.
- 33. Sarcasm
- 34. Sarcasm Can I help you?
- 35. Sarcasm Can I help you? No, I just waited in line for 30 minutes to say hi.
- 36. bAbI
- 37. bAbI Stories
- 40. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida.
- 41. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida. Where is John located?
- 44. Where is John located?
- 45. Where is John located? Word Sequence
- 46. Where is John located? Variable Length
- 47. Where is John located?
- 49. Where
- 50. Where is
- 51. Where is John
- 52. Where is John Parameter Sharing
- 53. Where is John Located
- 54. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida. Where is John located?
- 55. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida. Where is John located? Florida
- 56. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida. Where is Mary located?
- 57. Mary journeyed to Texas. John went to Maryland. Mary went to Iowa. John travelled to Florida. Where is Mary located? Florida