在嘗試勾勒影音串流技術發展的未來樣貌時，很難不去關注其它技術領域。影音技術徑自會有更多的發展是毋庸置疑的，但來自其它領域的突破與創新，無疑會帶來更多的可能性。第一個我想到的，就是人工智慧（AI）…

When considering the future of video streaming services and technologies, it is impossible to ignore the potential impact of new technologies from other fields. While video technologies themselves will continue to advance and deliver new value, the rapid pace of technological change means that the integration of these technologies with those from other fields could lead to even more opportunities in the future.

人工智慧肯定會在接下來的影音技術發展中，佔有一定的影響力。

舉個簡單例子：人工智慧可以更加優化影音串流時的網路傳輸。它可以根據所有人提供給它的資料，給與不同的頻寬使用建議，主動建議你的手機或電視接下來要下載的影音品質。而且，這還是它把你的各種情況給考量進來的情況 (無線網路品質、電信資費、剩餘電池量、個人偏好…等)。

再舉個例子：人工智慧還可以被用來分析你觀看影片的習慣，然後結合其它網路資源，進一步提供額外資訊給你。像是即將在三月開始的 F1 一級方程式賽車賽季，當你看到一半，想知道某車隊的一些過往歷史或是趣聞時，可以讓 AI 幫你準備好，然後看是要以虛擬主播的方式說給你聽，或是採用子母畫面的方式來撥給你看都行。

另一方面，AI 的再次成熟，讓分析影片，從裏頭看出一些端倪變得可以做到而且實務上可行。AI 將會開始看懂一些情境，像是「男主角第一次與女主角在哪相遇的？」「這部劇的主要拍攝景點是哪些？」「劇情急轉直下的關鍵是什麼？」「不同時間點的嫌疑犯各是誰？」…等，都有機會透過 AI 與人的合作下做到。

還有更多的發展樣貌，包括使用 AI 來最佳化 video encoding 等，都是已經正在進行式中的了～

Artificial intelligence (AI) is expected to play a significant role in shaping the future of video streaming. AI technologies, such as machine learning and natural language processing, can be used to improve the efficiency and effectiveness of video streaming services. For example, AI can be used to optimize the delivery of video streams based on network conditions, to recommend content to users based on their preferences, and to personalize the viewing experience. AI can also be used to analyze viewer behavior and provide insights to content creators and distributors, helping them to understand and engage their audience more effectively.

Extracting knowledge and information from videos has been a long-standing research area, with numerous studies and approaches developed over the years. For example, Schwenzow et al. [29] discusses the challenges and approaches for extracting knowledge from videos, including the use of natural language processing and machine learning techniques. Bai et al. [30] presents a survey of techniques for extracting information from videos, including the use of text and speech recognition, visual analysis, and multimodal approaches. These studies demonstrate the ongoing research efforts in this area, and highlight the potential for extracting knowledge and information from videos to deliver additional side-information.

references:

  • [29] J. Schwenzow, J. Hartmann, A. Schikowsky, and M. Heitmann. “Understanding videos at scale: How to extract insights for business research”. J Bus Res 123, 367–379 (2021).
  • [30] K. Bai R and P. Badar. “A Survey on Various Techniques used for Video Retrieval,” in PiCES, vol. 3, no. 1, pp. 15-17, May 2019.