日常前言
- 四月份花了一大部分時間去深入代碼，把我們的雙攝虛化流程解析了一遍。然後爲了給組內分享，又花了相當一部分時間去做總結，畫思維導圖、作流程圖等等，這其中學到了挺多東西的，尤其是對高通 Camera HAL 層的數據流部分，Camera Post Process 的前後節點都有了一個比較大概的瞭解，在跟蹤數據流的時候沒那麼頭暈了。
- 還有，總結、分享知識的時候，作圖真的非常重要，一份填滿大量文字的 PPT，可能講 3 個小時都講不完，最後聽衆也很難有所收穫。然而如果有 70% ~ 90% 的篇幅用圖片來直觀表述，剩下的文字用於精煉、簡潔地描述，這樣可能 1~2 個小時就能搞定，並且聽衆至少也能留下一個比較整體的印象。
- 好吧，扯遠了，迴歸這次的活動，這一期是物聯網的主題，又是我不熟悉的領域，只能找一些介紹性的文章來翻譯了。以及……上期又送來一個抱枕……這期要是再送公仔，那就把這些東西送給女盆友一宿舍當畢業禮物吧hh
- 這期採納了四篇：
版權相關
翻譯人：StoneDemo，該成員來自雲+社區翻譯社
原文鏈接：IoT in cognitive science and web proctoring
原文作者：（未找到作者信息）
IoT in cognitive science and web proctoring
題目：（物聯網中的認知科學與網絡監督）
IoT has been a buzz word for some time now and is being talked about in applying to all and sundry fields. But nowhere is there talk of applying it in cognitive science and web proctoring.
“物聯網（IoT，Internet of Things）” 這個詞已經流行了好一段時間，並且當前人們都在討論如何將它應用於各種領域。但是，沒有任何關於將其應用於認知科學（Cognitive science）和網絡監督（Web proctoring）的討論。
Cognitive science is all about measuring the effectiveness of a human in his cognitive endeavors. What we mean by cognition is related to how a human is learning, understanding and keeping new knowledge in memory.
認知科學，就是衡量一個人在其認知過程（Cognitive endeavors）中的有效性。我們所說的認知是指 “人類如何從記憶中學習、理解和保留新的知識”。
In massive open online courses (MOOC), students learn and take exams online. At present there is no way to properly proctor those online exams to be confident that the education imparted online is effective. Due to difficulties in effectively proctoring these online exams, there is a widely held perception that classroom-based education is somewhat more effective and desirable than online education. To change this perception, the effectiveness of proctoring has to be improved — this is where IoT comes into play.
在大型開放式網絡課程（MOOC）中，學生在線學習並參加考試。目前尚無方法對在線考試進行適當的監督，以確保在線教育是有效的。人們普遍認爲，由於難以有效監督這些在線考試，所以以課堂教學爲主的教育比網絡教育更有效、可取。爲了改變這種看法，監管的有效性必須有所改善 —— 而此處便是物聯網發揮其作用的所在。
I will explain how IoT can be an effective tool in cognitive science research and how IoT can help effectively proctor online courses in such a way to impart the badly needed credibility for online education.
接下來我將解釋物聯網如何成爲認知科學研究中的一種有效工具，以及物聯網如何以這種方式有效地幫助監督網絡課程，從而給予在線教育正急需的可信度。
Cognitive science and IoT
（物聯網與認知科學）
In cognitive science research, various sensors are fitted onto the subject and physiological parameters are measured and recorded. These parameters might be brainwaves measured through EEG or heartbeats, pulse rates, iris contraction, skin conductivity, etc. These parameters, when studied with other parameters which are related to how a person understood a certain topic or how the subject retained certain topic through a questionnaire, provide valuable insights. The sensors and sensor data are the subject of IoT wearables which can measure and send data through a wireless infrastructure to the cloud. If IoT technologies are adapted to the research of cognitive science, the data from various geographically different places can be collected, stored and studied with the help of the IoT cloud. Researchers will have valuable data for the asking and will actually “open source” the data to multiple stakeholders to conduct research effectively. At present, there is no such infrastructure which collects, stores and analyzes data from cognitive science related research.
A representative sensor pad for measuring EEG is shown below.
在認知科學的研究中，研究者將各種傳感器安裝到受試者身上，用以測量、記錄生理參數。這些參數可能是通過 EEG（Electroencephalo-graph，腦電圖） 或心跳，脈搏，虹膜收縮，皮膚電等方式測量的腦電波。在通過問卷調查來研究其它因素（關於人類如何理解一個特定話題，或者特定話題中的主題是如何保留下來的）時，前述的參數就能夠提供一些有價值的見解。傳感器和及其數據是物聯網可穿戴設備的主題，它可以通過無線基礎設施測量併發送數據到雲端。如果將物聯網技術適用於認知科學研究，則研究者就可以在物聯網雲的幫助下，收集、存儲並且研究來自不同地域的不同地點的數據。研究者們將獲得有價值的數據，並實際地 “開源” 數據給多個利益相關者進行有效的研究。目前，還沒有用於收集、存儲和分析認知科學相關研究數據的基礎設施。
下圖展示了一種具有代表性的，用於測量 EEG 的傳感器墊（Sensor pad）。
Unfortunately, the data from the headset stays with the nearest receptor and is stored in a local server. If the data is put into the IoT pipeline towards an IoT cloud, the data can be leveraged and properly “open sourced” for the greater good.
不幸的是，頭戴式耳機（Headset）中的數據與最最鄰近接收器保持一致，並存儲在本地服務器中。如果將數據放入物聯網管道（IoT Pipeline）中，實現物聯網雲，則數據可以被利用，並且能夠適當地 “開源” 以獲取更大的利益。
Where IoT infrastructure fits in is where data collection and dissemination has crossed certain limits. The headset shown above for EEG-based cognitive science research can be purchased off the shelf and used by anybody. This means with the availability of cheap off-the-shelf sensor pods, people who are interested in quantifying themselves will start using them. It would be a pity if no effort is made to collect the data, store and analyze it in a central cloud-based structure (which is nothing but an IoT cloud).
數據的收集與傳播已經存在某些侷限性的所在，就是物聯網基礎設施所適合的地方。上面展示的用於（基於腦電的）認知科學研究的耳機可以現貨購買，供任何人使用。這就意味着，有興趣量化自身的人將開始使用這些現成便宜的傳感器吊艙（Sensor pod）。如果不在收集數據上作出努力，並在基於雲的中央結構（物聯網雲）中進行存儲和分析，那就太遺憾了。
Where IoT fits in web proctoring and MOOC
（物聯網與網絡監管和 MOOC 相融合）
Another interesting scenario where the explosion of data occurs but is somehow ignored is web proctoring and massive open online courses. MOOC somehow lacks credibility when it comes to companies and employers accepting the certificates to be as credible as offline standard classroom course certificates. It is difficult to imagine a MOOC course conducted by Stanford to be as respected as the same course conducted in a classroom in Stanford. It all boils down to whether a student who took the course has sat for the exams in a manner in which the exams were properly proctored. In other words, it is difficult to proctor an online exam because the student is invisible and there are many ways to cheat in an online exam when compared to an offline one.
另一個有趣的場景是，發生了數據爆炸的情況，但卻不知何故被忽略了，這就是網絡監督和 MOOC 的情況。當涉及到到公司與僱主接受證書時，MOOC 由於某種原因而缺乏可信度（無法像線下標準課堂證書一般可信）。很難想象，斯坦福大學開設的 MOOC 課程能與其開設的線下課程一樣受到重視。這一切都歸結爲一個學習過該課程的學生是否已經參加了（以一種適當的監督方式進行）的考試。換句話說，由於學生是不可見（Invisible）的，並且比起線下考試，在線考試中有很多方法可以作弊，所以很難對在線考試進行監督。
Sure, there are platforms where students are monitored by online proctors through web cameras. Here a number of students’ webcams are monitored by online proctors, but there are no tools to capture other parameters of students who are taking the online tests. It is easy to cheat in an online exam just by keeping another screen somewhere in the room to get answers to the online questions. True, there are 360 degree cameras, but there is no data to validate that these are effective in curbing online cheating.
當然，有一些平臺可以通過使用網絡攝像頭（Web camera），使在線監考（Online proctor）可以對學生進行監控。這樣，許多學生的網絡攝像頭受到在線監考的監控，但沒有工具可以捕獲參加在線測試的學生的其他參數。僅僅需要在房間中另一個屏幕上查找在線問題的答案，就很容易在線上考試中作弊。沒錯，360 度相機是有的，但沒有數據可以證明它們在遏制在線作弊方面是有效的。
To end this menace and to increase the credibility of online courses and exams, a number of measures has to be taken. The most important one is to collect other kinds of data from the student along with webcam data, such as EEG, ECG, skin resistance and run analytics. One example is how the student reacts when he came across a questions which he feel is difficult. This can be achieved by carefully monitoring the above said sensors, namely EEG, ECG and skin resistance. Based on their inputs, the webcam data can be closely examined to see whether the student is trying to cheat or has already cheated, if the webcam data has been recorded.
爲了消除這種威脅，並提高在線課程和考試的可信度，人們必須採取一些措施。其中最重要的就是從學生那端，連同攝像頭數據一起，收集其他類型的數據（例如腦電圖，心電圖，皮膚電阻）並進行分析。舉個例子，當學生遇難題時，他是如何反應的。這可以通過仔細監測上述傳感器（即腦電圖，心電圖和皮膚電阻）來實現。根據這些輸入，可以仔細檢查攝像頭數據（如果該數據已經被記錄下來），以查看學生是否試圖作弊或已經作弊。
Interestingly, only the webcam data is enough to measure the heartrate, and this data will provide a certain amount of input by way of whether the pulse is increasing or decreasing while a student attempts to answer a question.
有趣的是，只有網絡攝像頭的數據便足以測量心率，並且這些數據將通過學生嘗試答題時脈搏的增減情況來提供一定數量的輸入。
The whole gamut of online exams and courses can be brought under the IoT umbrella and all analytics can be brought to IoT cloud, stored and analyzed for a more effective MOOC and online certification. Eventually more data can be brought in with the help of sensors available in any modern laptops namely mics for voice recognition, fingerprint sensors for identity management, facial recognition, etc.
所有的網絡考試和課程都可以歸入物聯網領域，所有分析都可以提供給物聯網雲，從而進行存儲和分析以獲得更有效的 MOOC 和在線認證。最終，藉助任何現代筆記本電腦所提供的傳感器（即用於語音識別的麥克風、用於身份識別管理的指紋傳感器，以及面部識別等），可以得到更多的數據。
So where does it all lead to?
（那麼，這一切會導致什麼結果呢？）
Based on the details about how IoT can be leveraged for both cognitive science and MOOC, a question may come about why cognitive science and MOOC are banded together in this article. The reason is obvious: Both fields are measuring physiological data in one way or another and can immensely benefit from IoT infrastructure and practices.
根據物聯網在認知科學和 MOOC 中如何得到充分利用的詳細說明，你可能要問了：爲什麼本文要將認知科學和 MOOC 結合在一起？原因很明顯：這兩個領域都以各種方式測量生理數據，並且可以從物聯網基礎設施和實踐中獲得極大的利益。
Therefore IoT can and most probably will help many fields mature and become widely available to the masses. This article highlighted just two fields which are poles apart but suited for IoT to take over because of commonality of data from physiological sensors. Their applications might be different, but the underlying platform can be same, to benefit all.
因此，物聯網能夠，而且很可能會幫助許多領域走向成熟，並廣泛地應用於羣衆。本文強調了兩個截然不同的領域，但是由於具有從生理傳感器中獲取數據的共性，他們都適合與物聯網相結合。他們的應用可能會有所不同，但底層平臺卻可以是相同的，並且都可以造福所有人。