幫朋友代po 網誌圖文好讀連結 https://medium.com/@chijuiwu/my-journey-to-chi-2017-e420dedc2cf5 --- My Journey to CHI술2017 CHI滹ﰱ7 Opening술KeynoteCHI滹ﰱ7 Opening Keynote (Panorama)Hi, My name is Chi-Jui Wu (喳‧ 蝖ꔠ ), or just Charles. I am an English teaching assistant in rural Hualien, Taiwan as a part of the s ubstitute military service program ( 蹂誨敶뤩. I am going to begin my Comput er Science PhD in Human-Computer Interaction at Lancaster University (UK), sup ervised by Dr. Steven Houben in the Interactive Systems research group, in Jan uary 2018. So I just got back from CHI 2017. The ACM CHI Conference on Human Factors in Computing Systems is the premier in ternational conference of Human-Computer Interaction. CHI滹ﰱ7 was held in Denver, Colorado, USA [~25% acceptance rate; 2424 submi ssions; ~3000 attendees]. In basketball terms, CHI would be like the NBA all-star weekend. I was extreme ly lucky to have 20 minutes on the court as a rookie滹 滹寣兩€ presenting my first-author paper EagleSense at my first academic/research conference. Man y people liked my work and my presentation. Therefore, I thought I should writ e a blog to reflect upon myself and my work leading up to CHI, and also share my conference experience. To really understand how I got my first CHI paper, I must look at my entire li fe experience as a whole. I will share my personal journey from being a shitty undergraduate student to getting an excellent PhD position to presenting a pa per at CHI. Perhaps you can滹槺 relate to my stories, but I will try to prov ide an unbiased account of my successes and failures, as well as strengths and weaknesses, so you may still benefit from the lessons I have learned along th e way. This blog post is not about how to write a CHI paper or how to to do re search (I am new too), but how I have developed personally and professionally to be able to present at CHI. I believe that only you could judge your own success, and everyone has differe nt histories, interests, motivations, goals, challenges, and triumphs, etc. Li fe is unpredictable, and our aspirations change overtime (I wanted to become a policeman when I was 7). Nonetheless, we should strive to become the best we can be滹 滹寣兩€ I try to be the best Charles I can be. Also, we should fi nd true happiness in life, a lesson which I will keep emphasizing throughout t his writing. My personal reflections may seem overly critical in places, but I want to keep it authentic to show how I really see myself in the past, presen t, and future. This blog post marks my new journey to becoming one of the top HCI researchers, and it will happen滹嶆Probably not, but I want to be bold. --- Being an Anti-Studying and Anti-Social Kid I had no idea what I was술doing. American International School of GuagnzhouI was born in Taiwan, and I moved to China at the age of 10. I graduated from American International School of Gua ngzhou, the best international school in the area, but it was super bad in so many aspects. I honestly think education is broken; it滹榊 a difficult probl em. Education should be about building a great human滹 滹寣兩€ intellectua lly and emotionally滹 滹寣兩€ and answering the question 滹嫷ho do I wan t to be?滹 (This idea is from Stanford University, which really does amazing work in education.) Before going to university, I never enjoyed studying and going to school. I li ked solving problems, and I knew getting good grades was the easiest way to le ave options open. I never planned on going to graduate school. I didn滹槺 kn ow what I wanted to do in life, so in that regard, I would say I almost failed high school (or high school failed me). I didn滹槺 have passion for anythin g. I was stressed at times, and I felt very lost. I guess I just kept going. I slept through lots of physics classes, but I enjoyed those extra sleeping ho urs. I actually did alright on my IB exam (5/7 score) after cramming the entir e textbook in one weekend, and I could now appreciate the beauty of physics. I wasted a lot of time playing computer games, and I wasn滹槺 even good at it (Defense of the Ancients). If I were to go back in time, I would try to play the game of DotA better. It滹榊 sad that I don滹槺 play games anymore. I had terrible social skills. Computer Science just seemed like a good fit. I was glad we had a course in In formation Technology in a Global Society. I spent a lot of time in front of co mputers, and I had a lot of patience for problem-solving (I still do). Looking back, I would definitely like to start programming sooner. I only started lea rning the art of programming when I got to St Andrews. Unfortunately, I think it was already too late to become one of the best. In high school, I noticed S ocial Science Professor Sherry Turkle at MIT from her TED talk. Since then, I started thinking about the relationship between humans and computers, and now I am doing HCI. It滹榊 still my dream to study at MIT. I was a hard-working and above-average high school student (3.71/4.00 GPA). Li ke many other Asian kids, I spent a summer cramming SAT, and I got an above-av erage score (1970/2400 score). It滹榊 retarded, but I had to play the game. Along with good recommendation letters and an over-hyped personal statement, I received an offer from the School of Computer Science at University of St And rews (UK). I am very grateful to come from a wealthy middle-class family and have somewha t decent English language skills (exactly because I come from a wealthy family who sent me to an international school), so I had an easier time abroad. I ha d more socioeconomic advantages than most people on Earth. --- Being a Shitty Undergraduate Student I worked a lot, a술lot. School of Computer Science, University of St술AndrewsI was rejected by many un iversities for undergraduate admission, including University College London (w here I later did my Master滹榊) if I remember correctly, but whatever. I thi nk I was meant to go to St Andrews, and I loved it. St Andrews is the best pla ce in the UK for undergraduate Computer Science, and I highly recommend high s chool students to go there. Four years were lost in the midst of attending/skipping lectures, programming/ hacking, being shit in team but winning hackathons, being mediocre at internsh ips, drinking, gaming, traveling, falling in love, getting heart-broken, and m ore programming/hacking. Although I was always in the lab (almost daily), so w ere my best friends, I still had a vague idea of what I truly wanted to do. So in that regard, I had once again almost failed higher education (or higher ed ucation failed me). I lost motivation, many times. I was depressed. I think at some point Computer Science students, or top university students in general, get symptoms of depression, because we try too hard. Luckily, I have always be en able to get through it by myself. I am really good at coping with immense s tress and sadness술:) I still had terrible social skills, but I made friends from all over the world , because I went to St Andrews. I had been really bad at finding interesting projects/topics to do/study. I th ink it had to do with my personality, which finds many things in life boring, and I don滹槺 care about a lot of things, too. I was very fortunate to have an extremely enthusiastic and encouraging thesis supervisor, Dr. David Harris- Birtill. Professor Aaron Quigley was also very keen on getting my work publish ed. I was passive, too. However, good things always happened when I took initiativ es. The summer before my final year, I interned at the St Andrews HCI lab (I w as given the opportunity because I did well in HCI course), and David was that very keen and positive postdoc. I had planned to do my undergraduate thesis i n game development with type-dependent programming language Idris, which was f ine but I wasn滹槺 passionate about it. I decided that I should ask David to be my thesis supervisor and work on a HCI project instead. It was one of best my decisions in life. David was always encouraging in weekly meetings, which invariably began with 滹嫷ell done, Charles!滹, no matter how little work I had done. I also met my future supervisor at UCL, Dr. Nicolai Marquardt, dur ing his seminal visit to St Andrews. He was incredible, and I learned so much from his talk. I wanted to do exactly the kind of research that he did. Proxem ic Interaction was WOW!! So I showed him a demo of my early work on Out of Sig ht and said hi. In my first few weeks at UCL, I contacted him about supervisin g my Master滹榊 thesis, then Nic introduced me to Steven. And last week, we presented our work on EagleSense at CHI滹ﰱ7. Most of the time I worked really hard, but often not cleverly. In retrospect, I was struggling in my first two years of undergraduate degree. For instance, I was very slow at translating ideas into code. I practiced a lot, but I was s till nowhere near the best. Again, I just kept going. It was normal for me to leave the lab at 3 or 4 in the morning and have endless all-nighters. In my fi nal year, I would often start working on my thesis at 11 p.m. and stay in the lab overnight, and my research progress suffered, but it was all for good roma ntic reasons. I sacrificed a lot of sleeping time and weekends to get shit don e. Like most people, I barely slept before major deadlines. I now remember Psy chology Professor Paul Gardner at St Andrews (I tried to go Computer Science a nd Psychology double-major) once told me that it滹榊 really bad to have a wo rking schedule like this and not be able to go to 9 a.m. lectures. Apparently, I didn滹槺 work hard enough to publish my work from research int ernships. Research is difficult, but I now think failure (to obtain results or publish) is just a part of the process. I lost motivation, but I was also mot ivated by failures and fed off broken code. However, I didn滹槺 滹廛ee滹 nor 滹廡nderstand滹 research when I was an undergraduate. I was just tryi ng to build cool stuff. At St Andrews, I was surrounded by groups of really smart and motivated indivi duals, whom helped me become a better person (also a programmer and a research er) in many subtle ways. I would not have accomplished so much without them. I acquired the skills and opportunities to eventually publish a CHI paper. I wo uld not have become who I am today if I went to another university (say, Unive rsity of Edinburgh). At last, I knew I enjoyed programming and doing research, but programming or researching what? I wrote another over-hyped personal stat ement, and along with good recommendation letters and grades (16.6/20.0 GPA, f irst-class), I was accepted by UCL, which offered what I vaguely wanted to do€ 滹寣兩€ machine learning and human-computer interaction. Also, my first CHI submission was brutally rejected (1.0/5.0 score). --- Being a Research Graduate술Student I also worked a lot, a술lot. Main Quad, University College술LondonI wanted to do research and publish. Natu rally, I spent most of my time working my research project. I really liked the course on reinforcement learning, which was taught by Google DeepMind researc hers. I attended many research talks, seminars, and workshops while I was in L ondon. I had to constantly keep myself motivated. It was the toughest year in my life, and I went through it alone, for the most part. Life was hard. I felt pressure from both academic and personal life. I had a lot of negative energy . I collapsed, both mentally and physically. In fact, I was close to failing my degree several times. I actually failed my reinforcement learning coursework, and I sat in my first exam (Graphical Model s) being completely blanked out and knowing I might not graduate. I just tried not to give up. Thankfully, I was able to negotiate grades with the marker. I gamed the system and finished with distinction. I struggled in research, too. EagleSense went through many iterations before i t actually works, and it still has many limitations. I read a lot of papers, s ucceeded and failed at implementing their methods. Our submission of the early prototype was straight rejected by another conference. Further, I didn滹槺 quite grasp the concepts of research and framing. I had only come to 滹廛ee€鴠research much later when I finished the EagleSense CHI paper and my PhD re search proposal. I almost gave up. I didn滹槺 believe EagleSense actually had a shot at CHI, well, it滹榊 CHI! And I had zero publication experience. Steven was kind of pissed at our paper not accepted to the other conference, and later he said Ea gleSense had a good chance at CHI. Either he really believed so or he lied to keep me motivated, anyway, I listened his wise words. I also didn滹槺 want t o give up yet, so I worked my ass off. Not really, I worked my ass off for a y ear already, so I was exhausted. Somehow I used all my remaining energy to imp rove the system and write the paper. I wanted to publish. London was a magical place, and I had a few but good friends. I also had a won derful trip to Spain before submitting my thesis, which was one of the few goo d memories during my Master滹榊. --- Getting a CHI술Paper It was술surreal. EagleSense CHI滹ﰱ7 paper술acceptedThe reviews were mixed but overall positi ve, and it was on the line of accept/reject. I had to write a strong rebuttal to push the EagleSense paper over the acceptance bar, but it turned out many p eople liked our work. So I guess rebuttals do work. I couldn滹槺 believe my paper got accepted. I didn滹槺 think I was any bet ter than other researchers in the CHI community, so I thought it must have bee n luck. I wasn滹槺 excited滹 滹寣兩€ I said to myself, 滹嫷hat滹榊 g oing on? Is it real?滹 I wasn滹槺 happy, too. I just felt more pressure, b ecause I thought my work was not at CHI滹榊 caliber. Therefore, I had a toug h time getting started on my CHI presentation. It was probably imposter syndro me. I was never confident about my work, because I always wanted to do better. If my paper were not accepted to CHI or any other top-tier conference, I woul d think I was not good enough to do research (especially since I had already s pent one year on it), and I would probably not have pursued a PhD. I had been unhappy in life, but CHI changed me. I was excited about going to CHI, and it was all uphill once I got to Denver, except the thunderstorm near my Airbnb ho use. I feel like my IQ must have boosted by 100 points after my paper got accepted. --- Getting a술PhD Good things술happen. Lancaster UniversityBefore I even started my Master滹榊 degree at UCL, I kne w I would be able to get a PhD position at St Andrews or UCL later on if I rea lly wanted. I was good enough. However, I was only absolutely certain about do ing a PhD after I graduated from UCL. I wanted to do my PhD in cross-device in teraction, but I also wanted to venture out, not going back to St Andrews or s taying at UCL. I was lucky to have find an excellent PhD position. Steven was going to Lancaster University to start his lectureship, and he need ed PhD students. He is an expert in cross-device interaction and is reasonably smart. Actually Steven surprises me every time I talk to him; he has a lot in his mind. I was a top student at UCL, and we wrote a CHI paper together. Stev en wanted me to go to Lancaster, and so do I. I was accepted to the substitute military service program in Taiwan (military service is still compulsory here), and I got a really good teaching position w ith many interesting teaching challenges. I had enough free time to write my P hD research proposal and prepare my CHI talk. I was also able to attend CHI滹™17 in Denver, which would otherwise be impossible. This blog post is special ly dedicated to the students at Wanrong Junior High School. They have so much potential!! As their teacher, I really want them to do better than me in life. I applied only to Lancaster University for PhD. I got it. The fellowship inter view was rather intense, but I liked the questions from my reviewers. One of t he tough questions was research versus engineering, and they kept challenging my responses. If they didn滹槺 make me feel this way, I would probably have been disappointed. And Steven said it was good, since the reviewers would only ask difficult questions if they were really interested in you and your work. I got the fellowship. I worked a lot during my undergraduate and master, and often I was the last on e to leave the lab (if I ever decided to stay). I would probably do the same a t Lancaster. --- Attending CHI滹ﰱ7 CHI滹ﰱ7 was fun, amazing, inspiring, and life-changing. I was too nervous. I need to remember to take a picture of the audience from m y field of view next time (maybe UIST滹ﰱ8술?)CHI is a huge networking event . I traveled to CHI滹ﰱ7 alone, but I met many other students, professors, a cademic and industry researchers at the conference and parties. It found it di fficult to socialize, especially with senior researchers (so pretty much every one except students), but I think I did alright for my first conference. I cou ldn滹槺 participate in the Student Volunteer program this year (because of t he military service), but I滹朞 definitely like to get involved at CHI滹ﰱ 8. It was so cool to finally see the top researchers and HCI pioneers in real life, and they were all very nice. I also got to know more about the CHI commu nity from Taiwan. I wanted to talk to Dr. Pei-Yu Chi from Google, who worked o n Weave and DemoScript, both really good work, but unfortunately it didn滹槺 happen술:( I actually liked the breadth of research covered at CHI, but it was also a ver y intense schedule. I had about 4 to 6 hours of sleep each day, because I was jet-lagged and nervous about my presentation, so I woke up early in the mornin g to practice my presentation. Hence, I kept falling asleep in the afternoon s essions. Naturally, I was exhausted by the last day of the conference. My talk went really well, as I was told, but I just tried not to get too nervo us on stage while talking to the microphone. Less than 50? people showed up to the camera-based tracking session, but I still got good positive feedback aft er my talk, and I would like to incorporate their ideas into our follow-up wor k. Andy Wilson from Microsoft Research suggested a paper that I could use to e xtend the EagleSense tracking infrastructure. I really liked other work presen ted at our session, and I thought the short skits shown in Dr. Jens Grubert滹™s talk were on point and hilarious. I don滹槺 have a favorite talk at CHI€™17 because there were so many (pen-touch interaction, cross-device interac tion, sensing techniques, case studies, the physical web, room-scale collabora tion, social responsibilities, Professor Jacob Wobbrock滹榊 social impact aw ard talk,술...). The keynotes and panels were fantastic. I will list a few things that I didn滹槺 like about CHI滹ﰱ7 before saying how wonderful it is again. 1) The panel after the opening keynote was awkward . There was hardly any discussion. 2) The closing keynote was plain and uninsp iring. 3) Too many parallel sessions. 4) No video recording for some sessions. I am just sad that my talk was not recorded. 5) Lack of conference gifts, not even a CHI滹ﰱ7 pen or notebook. Also unimaginative merchandises. The opening keynote was WOW!! Neri Oxman discussed how us humans can coexist w ith nature through novel designs in the Material Ecology. She has a brilliant mind. Despite talking about her amazing work, she reiterated the importance of societal contributions and impacts. What history/legacy/impact will the Eagle Sense paper leave, if any? What contributions do I really want to make? And is it even possible to make a profound impact during my three-years PhD? I think real impact is only ever so substantial and tangible when looking at the brea dth of a person滹榊 work over his lifetime, such as Ben Shneiderman滹榊. 滹嫿he best is yet to come!滹滹 滹寣兩€ Ben Shneiderman 滹嫿he best way to predict the future is to invent it.滹滹 滹寣兩€ Ala n술Kay I truly believe in these words. I have so many research ideas after attending CHI, and it滹榊 really up to us to fulfill our own research visions. I want to bring cross-device interaction to novel and important application domains, including education, healthcare, and economic development, empowering individu als (inclusiveness and accessibility) and enabling striving communities. I don 滹槺 know how (and perhaps not through cross-device interaction research), b ut I strongly believe that doing research in this direction can really help th e world become a better place. And on this note, looking back at all my previo us work and my most recent PhD research proposal, I have not been bold enough even in my wildest visions. Lastly, I want to talk about a special friend I met at CHI. I always believed that we could meet again at CHI, and we did! Although we had vastly different skill sets (one is good at design; the other is good at implementation), we bo th wanted to do great HCI research. I was truly happy to see us en route to ac hieving our dreams despite taking different paths. I would never forget all th e memories we shared. It was the best thing that had happened to me at CHI滹™17. --- Life Lessons I learned a술lot. Pursue what really matters to you and the kind of person you want to become. K nowing what I really I wanted to achieve kept me motivated in times of utter d espair. Nothing is easy. If the thing you are working on is easy, you are not learning and improving. Or you are a genius. More hacking, less optimizing. There is a trade-off between hacking new things and writing beautiful code. The best are good at both. Meet different people. Be proactive. I always learn a lot from other people. C reativity requires exploration. At CHI, everyone was interested in different t hings, and by talking to them, I formed new ideas for my own research. Have fun and be truly happy. I actually have no idea how to have fun, and it€™s really difficult for me to be truly happy. I am always chasing the thing s I don滹槺 have, be it new knowledge or a CHI paper, and I am never satisfi ed. I think achieving work-life balance will become one of my most important t opics in life, and it will determine how much success I will have. Thank you for everything. --- I have always fought the battle alone, but I always have support from my famil y, friends, and supervisors. I just work really hard, and people like my work. You need to trust yourself. Trust the process. In the end you will achieve yo ur goals. Just keep doing research, and you will succeed. You will fail someti mes, but they are just really high walls to test how much you want the thing y ou are chasing (see also Randy Pausch滹榊 Last Lecture). One day you will ma ke a breakthrough. You will write a CHI paper about it, and get rejected. You will write another, one after another, and eventually you will get a CHI paper accepted, an honorable mention, and a best paper award. But it滹榊 never ab out the papers. It滹榊 just a part of the process. It滹榊 also not about t he research. It滹榊 about making the changes you want to see in the술world. --

※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 223.137.162.202 ※ 文章網址: https://webptt.com/m.aspx?n=bbs/HCI/M.1495202134.A.C34.html