1. Information about the Paper


Parate, Abhinav, Meng-Chieh Chiu, Chaniel Chadowitz, Deepak Ganesan, and Evangelos Kalogerakis. “RisQ: recognizing smoking gestures with inertial sensors on a wristband.” In Proceedings of the 12th annual international conference on Mobile systems, applications, and services, pp. 149-161. ACM, 2014.


Smoking-induced diseases are known to be the leading cause of death in the United States. In this work, we design RisQ, a mobile solution that leverages a wristband containing a 9- axis inertial measurement unit to capture changes in the orientation of a person’s arm, and a machine learning pipeline that processes this data to accurately detect smoking gestures and sessions in real-time. Our key innovations are fourfold: a) an arm trajectory-based method that extracts candidate hand-to-mouth gestures, b) a set of trajectory-based features to distinguish smoking gestures from confounding gestures including eating and drinking, c) a probabilistic model that analyzes sequences of hand-to-mouth gestures and infers which gestures are part of individual smoking sessions, and d) a method that leverages multiple IMUs placed on a person’s body together with 3D animation of a person’s arm to reduce burden of self-reports for labeled data collection. Our experiments show that our gesture recognition algorithm can detect smoking gestures with high accuracy (95.7%), precision (91%) and recall (81%). We also report a user study that demonstrates that we can accurately detect the number of smoking sessions with very few false positives over the period of a day, and that we can reliably extract the beginning and end of smoking session periods.


2. My review of the Paper


The paper proposed a better way to recognize smoking gesture utilizing a wristband with 9-axis inertial sensor. The paper successfully articulated the method and proved/demonstrated evaluation of its accuracy (95.7%). The paper also compared its method with previous work eloquently. This method can also be altered to detect eating gestures.

Its key contributions are:

  1. extracting hand to mouth gesture using arm trajectory-based method
  2. it has trajectory-based features to distinguish between smoking and similar hand to mouth gestures such as eating and drinking.
  3. it can infer smoking session from these data
  4. reduce burden for users to self report those data



  1. The paper is more well written, compared to previous papers reviewed in class, although has few grammar and spelling error.
  2. The paper clearly stated previous related work and its contribution/advancement from previous work.
  3. It is better than previous work in term of accuracy and practicality of using it in real-world application. Because previous works used sensors that were not easy to wear daily.
  4. Previous work such as BiteCounter request user to push a button to detect a session start. Whereas RisQ does not request this information. RisQ did request this information, but only for accuracy evalution. While BiteCounter used this information directly in their algorithm.


  1. The paper seems to not mention any limitation.
  2. There is a simple numbering error where Figure 10 is presented before Figure 9.
  3. The paper did not mention or compare using barometer as a sensor. I think barometer could be useful to detect smoking gestures. Assuming that barometer is worn on the users wrist, it should detect the height difference of the wrist in a smoking session gestures easily.
  4. The paper did not mention this closely related work from 2012 UbiComp, “A Feasibility Study of Wrist-Worn Accelerometer Based Detection of Smoking Habits” which has a really close idea with this paper. The paper should compare itself to this work. This work used only one axis accelerometer on the wrist and simple classifier and did successful in detecting smoking gestures. Although with lower accuracy.
  5. The paper did not mention demography (gender/sex) of the participants which could be useful for more specialized interpretation.

Related/Other Comments:

  1. It has not been cited by any following work, because it is very recently published (2014). Does not mean it is not significant.
  2. It might help smokers that has an intent to stop smoking gradually. But it won’t help smokers who has no intent to quit smoking. We need non technical persuasion for this.