Cognition · conference · Haptics · Multimodality

Preparing submissions for the SweCog 2017 conference, held at Uppsala University!

SweCog2017_Uppsala

This week, I’m preparing submissions for this year’s version of the SweCog (Swedish Cognitive Science Society) conference. This conference covers a broad range of topics related to cognitive science. When I participated last year, when the conference was held at Chalmers, Gothenburg, I did not present anything (actually, none of the participants from Uppsala University did), but the situation this year is quite different since Uppsala University is hosting the event!

I really enjoyed last year’s conference much due to the large variety of topics covered and the very interesting keynote lectures. It was also (and still is, I assume) a single track conference, meaning that you will not have to choose which paper session to attend. As I remember there were ten paper presentations in total, three keynote lectures and one poster session during the two days conference. You can read more about my experiences from SweCog 2016 in this blog post, summing up that event. I also wrote summaries from day 1 and day 2.

Since the only thing that’s required is an extended abstract of 1-3 pages (and max 500 words), I’m working on several submissions. A topic that was not covered during last year’s conference was collaboration in multimodal environments and specifically how different combinations of modalities can affect communication between two users solving a task together. Since that is one of my main research interests, I now see my chance to contribute! The deadline for extended abstract submissions to SweCog 2017 is September 4, so there is still a lot of time to write. The conference will be held October 26-27 at Uppsala University. Since registration to the conference is free for SweCog members (membership is also free), I expect to see many of my old KTH colleagues at Uppsala University during the conference days! 😉  You can find more information about the conference here.

Before I started planning for contributions to SweCog 2017, I invited some of my “multimodal colleagues” from KTH to join the writing process. As a result, Emma Frid and I will collaborate on an extended abstract about a follow-up study to the study I present here. Thus, our contribution will focus on how multimodal feedback can affect visual focus when two users are solving a task together in a collaborative virtual environment. Since I have not yet heard from any other colleague, I plan to write another extended abstract on my own, about how multimodal feedback (or rather combinations of visual, haptic and auditory feedback) can affect the means by which users talk to each other while working in collaborative virtual environments. Maybe, I will also throw in a third one about the potential of using haptic guiding functions (see this blog post for an explanation of this concept) in situations where sighted and visually impaired users collaborate.

 

conference · Haptics · Multimodality · sonification

Got a new paper published, on the effects of auditory and haptic feedback on gaze behaviour!

SMC_published

About a month ago I wrote a blog post about a conference paper with the title “AN EXPLORATORY STUDY ON THE EFFECT OF AUDITORY FEEDBACK ON GAZE BEHAVIOR IN A VIRTUAL THROWING TASK WITH AND WITHOUT HAPTIC FEEDBACK” that had just been accepted for the Sound and Music Computing 2017 conference. Now, that paper has been formally published! You can find our paper here and the full conference proceedings here. The study leader, Emma Frid presented the paper last Thursday (6/7 2017) afternoon in Espoo, Finland. The other authors are Roberto Bresin, Eva-Lotta Sallnäs Pysander and I.

As I wrote in the earlier blog post, this particular paper is based on a small part of an extensive experiment. The experiment, which 20 participants took part in, was based on a simple task – picking up a ball and throwing it into a goal area at the opposite side of a virtual room. After 15 hits the task had been solved. The same task was solved in several different conditions of which some included haptic rendering and some included movement sonification (two different sound models were compared) of the throwing gesture. During all interaction with the interface, different parameters, including gaze data collected through an eye-tracker, were continuously logged. In the part of the experiment on which the published paper is based we wanted to find out if the participants’ visual focus in the interface changed depending on experiment condition (e.g. if participants looked more at the goal when haptic and/or auditory feedback was presented). Due to bad quality of the sampled gaze data for some of the participants (< 80% of the gaze points had been registered), only gaze data from 13 participants could be used in the analysis.

Much due to large inter-subject variability, we did not get any significant results this time around, but some interesting patterns arose. Results e.g. indicated that participants fixated fewer times on the screen when solving the task in visual/audio conditions compared to a visual-only condition and fewer times on the screen when solving the task in the visual/haptic/audio conditions than when doing it in the visual/haptic condition. The differences between haptic conditions were, however, small especially regarding one of the sound models presenting a swishing sonification of the throwing gesture. When considering total fixation duration (for how long the participants focused on the screen) the tendency was that participants focused less on the screen when this sound model was used (indications were stronger when haptic feedback was not provided). Even though these results were not significant they indicate that movement sonification has an effect on gaze behaviour. When looking at gaze behaviour for each participant individually we could also see that the participants could be divided into a few clusters in which the participants showed similar behaviour. Although the large inter-subject variability did not make it possible to find any general patterns, we could find indications of effects of auditory feedback within the clusters. See the article linked above, for a more detailed analysis, illustrations and discussion.

Even though we did not get any significant results, the indications we got that movement sonifications can affect visual focus are still interesting. If it is true that you look more on the screen when you do not have access to movement sonification, this can mean that you can focus on different parts of an interface, maybe solving different tasks in parallel, when having access to movement sonification in this kind of environment. It is definitely worth conducting similar studies with a lot more participants in order to see if the indications we got would become significant. Experiments with more users could also show if participants focus more on the goal when having access to movement sonification and/or haptic feedback – if so, this would indicate that the information provided by haptic and audio feedback, respectively, is enough to understand that you are performing an accurate throwing gesture (you don’t need to look at the ball to confirm it). Results from interviews held at the end of the test sessions already indicate this!

This is the very first paper Eva-Lotta and I have gotten accepted to the Sound and Music Computing conference. Emma and Roberto, however, have gotten papers accepted to that conference numerous times. Check out their Researchgate accounts for their earlier contributions to this conference and so much more related to e.g. sound design.

games · Haptics · Multimodality · sonification

Paper on the effect of auditory feedback on gaze behaviour accepted to SMC 2017!

SMC_accept

Earlier this week I wrote about a paper that was accepted to the Frontiers in Education (FIE) 2017 conference, but the fact is that yet another paper which I co-authored was accepted to another conference, Sound and Music Computing (SMC) 2017, earlier in May! Emma Frid (lead author), Roberto Bresin and Eva-Lotta Sallnäs Pysander from the department of Media technology and Interaction Design at the Royal Institute of Technology (KTH) are the other authors on that paper. The title of the SMC paper is “AN EXPLORATORY STUDY ON THE EFFECT OF AUDITORY FEEDBACK ON GAZE BEHAVIOR IN A VIRTUAL THROWING TASK WITH AND WITHOUT HAPTIC FEEDBACK”.

The paper is based on a small part of an extensive study, focusing on the effect of haptic and audio feedback on perception of object qualities and visual focus, performed a few years ago. In this particular paper we use eye-tracking metrics to investigate if auditory feedback in particular affects gaze behaviour in an environment where the task is to pick up a ball and throw it into a target area. We looked at both the effect of sound in general and effects of different sound models. Like in many other studies we have been involved in, conditions with different modality combinations were compared against each other. I will write more about the results when the paper has been presented and there is a link to the published proceedings. Search for the title given above if you want to find the specific session and listen to Emma’s presentation at the conference!

Here is the abstract, summarizing the main points:

This paper presents findings from an exploratory study on the effect of auditory feedback on gaze behavior. A total of 20 participants took part in an experiment where the task was to throw a virtual ball into a goal in different conditions: visual only, audiovisual, visuohaptic and audiovisuohaptic. Two different sound models were compared in the audio conditions. Analysis of eye tracking metrics indicated large inter-subject variability; difference between subjects was greater than difference between feedback conditions. No significant effect of condition could be observed, but clusters of similar behaviors were identified. Some of the participants’ gaze behaviors appeared to have been affected by the presence of auditory feedback, but the effect of sound model was not consistent across subjects. We discuss individual behaviors and illustrate gaze behavior through sonification of gaze trajectories. Findings from this study raise intriguing questions that motivate future large-scale studies on the effect of auditory feedback on gaze behavior.

As was the case with the FIE paper mentioned earlier, the SMC paper is just presenting a small part of a large study, so there is definitely a lot more to tell about the study and the different parameters measured. I will return to the overall study as soon as more papers are out!  🙂

 

communication · Group work · Haptics

More about my work with haptic communicative functions in collaborative virtual environments

img_1120

I’m currently in the process of writing a job application for an associate professorship and to make sure I don’t miss anything I recently browsed through my old conference folders to find the articles to append. When I came to EuroHaptics 2010 I was reminded about something I had completely forgotten – I was actually one of the organizers behind a workshop at that conference! I spent quite a lot of time preparing for the workshop, which focused on haptic communicative functions, but the weekend before the conference I got very sick and was forced to cancel my participation. I will take this opportunity to briefly introduce the workshop and discuss some of my work prior to it. My earlier blog posts on haptic feedback as an interaction modality were the following:

It was a shame that I could not attend the workshop/conference in Amsterdam, since it was based on my work on collaboration in multimodal environments up to that point. Thus, this was the perfect opportunity to discuss the work performed and get input regarding future work in the area. We described the focus of the workshop in the following way:

In this workshop, concrete examples will be presented and discussed in terms of how the touch modality can support communication and collaboration. Also, the technical challenges of distributed haptic feedback will be addressed. The target audience of the workshop is researchers and practitioners focusing on haptic feedback supporting people in settings where more than one user are involved. We invite other researchers and practitioners to share their research and experience from their different projects focussing specifically on the collaborative perspective. It might be that the collaborative aspects in your project have not yet been addressed. In that case, interesting collaborative aspects can be identified during the discussions in this workshop.

Quite a lot of work was performed by me and my “multimodal colleagues” 🙂 prior to the workshop. First of all, I had performed my master’s thesis work back in 2006 which focused on collaboration between visually impaired and sighted pupils in elementary school. Evaluations were performed in schools, where visually impaired and sighted pupils collaborated in dynamic collaborative environments where objects could be moved. During that work, and especially during a re-analyses performed during my first year as a Ph.D. student (2008) I realized that communicative functions based on haptic feedback had a real potential both when it came to supporting collaborative work and supporting inclusion of the visually impaired pupils in group work with sighted peers. It became especially clear that haptic functions for guiding (holding on to the same object or holding on to the peer’s proxy) can replace verbal guidance to a large extent.

Imagine a situation where you need to guide a visually impaired pupil to a particular place in a virtual environment. If you only have the visual feedback to rely on when establishing a common frame of reference, you need to talk a lot like  “go down, more down, …no, too much, go back…, now to the right…no, not down, up again… here it is!”. If you have haptic feedback available you can just grab the other person’s proxy and move the visually impaired peer to the right place and just say “here”. Needless to say, haptic feedback affects the dialogue between collaborators in this case. If you want to learn more about this explorative study you can read the journal article we finalized a few years after the study.

One problem that was evident from the evaluations with the visually impaired and sighted pupils was that the visually impaired pupil was not aware about what the sighted pupil did when haptic guiding was not utilized. This is why we performed a follow-up study where we added sound cues to the above mentioned dynamic interface to provide feedback on actions taken in the interface (e.g. grasping and putting down objects). We compared the new visual/haptic/audio version to the original visual/haptic one. We managed to show that the dialogue between the collaborators differed depending on which program version they worked in and it was also clear that the work was more effective (measured as time to complete task) in the visual/haptic/audio version. Once again, we could also clearly see how access to haptic feedback influenced the communication. You can read more about this study in this article.

These two studies resulted in a theoretical conference paper presented at HAID (Haptic and Audio Interaction Design) 2009, where we tried to develop a kind of conceptual model regarding haptic communicative functions’ effects on the dialogue between collaborators. This was my first attempt at a meta-analysis of the insights gained within this research area. The paper summarizes and discusses all the effects on the dialogue I had seen in my studies thus far. The paper made quite an impression – it is still the most referenced of all the papers I have produced up until today! At that point we were still quite unique when it came to haptic collaborative environments and I still think I’m one of the very few researchers who study the effect of haptic and audio feedback on the dialogue between collaborators.

The HAID conference paper laid the ground work for the workshop described in the beginning of this post and during the workshop the idea to study collaboration between two sighted persons was introduced and discussed. Next time I write about my earlier work I will introduce my latest study on collaborative multimodal interfaces, that showed that haptic and audio feedback indeed have effects on the dialogue between sighted persons as well!

 

deafblindness · Grant application · Haptics · sonification

Recently submitted my first ever research grant application!

One thing I have not mentioned in this blog before is that I’m one of the researchers behind the newly started network “Nordic Network on ICT and Disability”. This network gathers researchers from universities in Sweden and Denmark, focusing specifically on technology support for people with deafblindness. The reason why I’m a part of the network is primarily that I have developed some multimodal interfaces (based on haptic and audio feedback) for collaboration between sighted and visually impaired pupils in primary school (you can read this article and this conference proceeding for a summary of that work).

I have been thinking about writing a research grant proposal with a group of researchers belonging to the above mentioned network, ever since I joined it. And this year it finally happened! 😀  During a grant club in the middle of March, where several researchers from my division at Uppsala University gathered to write research grant proposals for a day, I ended up with a draft which felt close enough (read more about the very well organized grant club here). The draft was used as basis for discussion in a Skype-meeting with some other members of the network, after which we finally ended up with a research plan everyone felt comfortable with. It was submitted to the Swedish Research Council. I wrote most of the text, but it would never have worked without all the valuable input I got from my colleagues (most of them also co-applicants) both in the form of comments and addition of text chunks.

The proposed research focuses mainly on haptic feedback and how it can be used to support pupils with deafblindness in collaboration with sighted pupils – thus the focus is quite close to the research with visually impaired pupils which I, and several of the other co-applicants, were working on before.

The co-applicants are:

Apart from the above mentioned co-applicants, Charlotte Magnusson ([Research gate]) is also a part of the proposed project as a resource person from CERTEC.

I really believe in this team, since we complement each other in a very good way and we also belong to universities in Sweden which are in the top regarding research on assistive technologies and collaborative haptics. We of course hope the project will be funded, but in case it is not I really hope this team gathers again in search for other possible grants!

communication · Haptics

Haptic communicative functions

thesis

This is the fifth post in my blog series about haptics as an interaction modality and this time I start focusing on my main area – collaboration in multimodal interfaces. In earlier posts I have written about:

Ever since I started working with my master’s thesis project back in 2005 I have been focusing on collaboration in multimodal interfaces and specifically on how one can design haptic functions for collaboration and communication between two users working in the same environment. Collaborative haptic interfaces has been around for quite some time – one of the first examples is an arm-wrestling system using specialized hardware enabling two users to arm wrestle over a distance. Other commonly known, early, examples enabling a kind of mediated social touch are HandJive and InTouch. One thing these early systems have in common is that they use specialized hardware which is quite limited in scope. More examples of early systems and in-depth discussions about mediated social touch can be found in this excellent review.

During the two recent decades more widely applicable collaborative haptic functions for collaboration in virtual environments have been developed. Such functions can e.g. enable two users to feel each other’s forces on jointly held objects or make it possible to “shake hands” by utilizing magnetic forces between the two users’ proxies in the virtual environment. One of the earliest examples of an environment supporting these kinds of collaborative haptic functions, or guiding functions as I use to call them, is the collaborative text editor developed by Oakley et al.. Apart from the obvious functions needed to edit a document each user could also use a Phantom device to find positions of and/or communicate with the co-authors. An example of a haptic function was a kind grabbing function (similar to the shake hands function mentioned above), making it possible to grab another user’s proxy moving it to another part of the document. Other examples were a kind of “locate” function dragging one’s own proxy to another user’s proxy by a constant force or a “come here” function dragging another user’s proxy to one’s own position.

Later examples of virtual environments enabling these kinds of haptic collaborative functions are a collaborative drawing application developed at CERTEC, Lund University and applications for joint handling of objects evaluated by my former Ph.D. supervisor Eva-Lotta Sallnäs Pysander during her years as a Ph.D. student (thesis link). The latter examples are most relevant for me, since much of my work during my period as a Ph.D. student at KTH focused on collaborative interfaces in which two users can work together to move and place virtual objects. I will get back to my own applications later on, in a blog post about haptic interfaces supporting collaboration between visually impaired and sighted persons.

Above, I have just provided a few examples of collaborative haptic functions which can be used to control the forces provided by one or more haptic devices. The functions I think are most interesting to explore are the ones that enable physical interaction between two haptic devices in that both users can feel each other’s forces on jointly held objects or feel each other’s forces when holding on to each other’s proxies. These kinds of functions enable interesting means of communicating physically in virtual environments, especially in cases in which the users are not able to talk to each other face-to-face or point on the screen. Imagine, e.g. a scenario in which two users are exploring different parts of a complex multimodal interface showing distributed data clusters (what those clusters represent is not of importance here). In such an interface it would be very cumbersome to try to describe to the other person where a certain interesting cluster has been located. In this case the user who wants to show something s(he) found can grab the other user’s proxy and drag him/her to the relevant cluster. This possibility would probably simplify communication about the explored dataset (explaining where you can find details in a complex interface can be extremely cumbersome). This is, of course, just a made up example but it can be applied to many scenarios especially in cases where important parts of an interface are visually occluded. I will get back to joint handling of objects in a later blog post in this series. I will discussion the potential of using haptic feedback when exploring huge datasets in one of the upcoming blog posts.

Last, it is interesting to contrast functions enabling physical interaction between two haptic devices with functions only enabling a one-way communication (like the “goto”-function mentioned above). Using a one-way function enable some kind of communication in that one person’s proxy is “dragged” to another one’s, but the haptic function is only applied to one of the users in this case – there is e.g. no way for the other user to tell if the one being dragged actually wants to be dragged. When using a two-way haptic communicative function both users can feel forces from each other enabling a richer communication. Apart from enabling joint handling of objects, where the coordination of movement is made possible by both users feeling the other one’s forces on the jointly held object, two-way haptic communicative functions make it possible to e.g. clearly communicate to the other user that you do not want to be dragged somewhere. The potential these functions can have in situations where visually impaired and sighted users collaborate in virtual environments will be the topic of my next post in this series!

 

games · Haptics

Haptic feedback in games

falcon

Now it’s time for the forth post in my blog series about haptics as an interaction modality. In this post, I will write about games – an area where I think haptic feedback can be used in a much greater extent than it is today. The earlier posts in this blog series were:

Haptic feedback has been used in games for quite some time. I think that everyone has some kind of relation to the joysticks used in e.g. flight or car simulators. Most joysticks do not only enable some kind of steering, but also generate haptic feedback often in the form of vibrations or resistance to motion. If we take an ordinary flight simulator joystick as an example, the player can experience heavy vibrations when the plane is stalling, as a kind of warning that the lift is beginning to decrease.

During recent years new input devices have been developed with the potential to really change the way we experience different kinds of games. I have already introduced the Phantom Omni in earlier posts – a device that makes it possible to not only feel texture, stiffness, friction, etc., but also to lift and move around virtual objects. This clearly opens up new possibilities for game development, especially since the Novint Falcon (picture above) started to spread. As far as I can understand haptic feedback is, in the vast majority of games where this kind of feedback is utilized, still limited to vibrations and resisting forces despite the fact that modern devices greatly widen the possibilities. Below, I will add a few thoughts about what can be done to utilize the unique aspects of haptic feedback in games. There are, of course, many more things you can do apart from the ones discussed here.

Imagine, e.g. a haptic game where the player not only has to worry about navigating to the right place and/or interacting with different objects, but also need to watch out for deadly magnetic wells “physically” pulling the game avatar towards them. That would certainly add a unique dimension to a game, as would magnetic “guides” pulling the user in a certain direction making him/her aware that e.g. an object is approaching. Every year students are creating simple games, based on magnetic objects which should be avoided, in the haptics course at KTH. Here is an example video from a simple game where the user need to navigate through a mine field to find a treasure! It is easy to add more levels and objects, so the game is quite scalable and the idea can be applied to many different scenarios. Another game from another course round used a similar idea – that you should avoid being dragged into objects – but in that case the objects had different widths and were moving from right to left. The user should stay clear of the objects for as long as possible.

There are many games out there today which are based on the exploration of large and different environments. Zelda and the Final Fantasy series are among the most known examples. In those kinds of games haptic feedback could also add an interesting dimension, when it comes to categorizing objects and/or explore occluded areas hidden behind or within buildings, trees or cliffs. In these kinds of games you still need ordinary input controllers, of course, but a haptic device could be used as a complement. Imagine that you walk around in a large virtual environment and come to a well which you cannot go down into. You could then switch to a haptic mode and send down a probe to feel what is at the bottom. If something is down there you could also pick it up. You could even take this further and have small puzzles in hidden places (like in the well example), where you need to feel differences between e.g. friction, surface texture and/or weight of different objects. If you place the objects in the correct order you could unlock some secret.

Haptic feedback could also be used a lot more in puzzle and maze games – there are quite a few of them out there today. If you add a haptic feedback dimension to a puzzle game you can e.g. use weight and texture of different pieces as additional input. A haptic-only puzzle would be very interesting to try out! You can also play around with haptic mazes and use friction, texture and maybe even magnetic forces to provide additional information about where you are, provided that you cannot see your own location. Quite a few projects in the haptics course have been based on haptic mazes.

Above, I have sketched on a few ideas on how one can utilize some unique aspects of haptic feedback in games. Since we already have the technology, I think it is important that we try to take a step further from games where haptic feedback is limited to vibration, resistance and indications of getting shot at and instead look at more creative ways to use haptic feedback. There are some creative solutions out there today, but I think many games could still benefit from using e.g. the ideas discussed above!