Citation Context

...machines because it is computationally costly. In order to maximize the positive inference of an individual’s activity, prior work suggests that the best place on the body to carry a phone is the hip =-=[34]-=-. After interviewing the participants in our user study, we conjecture that most of people carry their phones in their pants pockets, clipped to a belt or in a bag. We collected training data from ten...

Citation Context

...e of physical activities such as walking, running, and resistance training is to wear multiple accelerometers simultaneously on different parts of the body (e.g., wrist, ankle, thigh, elbow, and hip) =-=[2]-=-. While this approach is known for yielding strong accuracy rates, it is not a practical form factor when considering all-day, everyday use. Another approach uses multiple types of sensors (e.g., acce...

Citation Context

... of the research using wearable sensors for activity recognition has so far focused on using a single sensor modality, typically accelerometers, placed in two or more (up to 12) locations on the body =-=[4, 5, 16]-=-. Placing sensors in multiple predefined locations or even a single, fixed, location every time can be quite cumbersome when one has to collect data on a daily and continuous basis. Work by [5] has sh...

Citation Context

...] [22] and urban sensing [11] [10] [1] applications. Although early work [36] mentions the use of microphone as a clue to context, most work found in the literature focus on the camera, accelerometer =-=[6]-=-, and GPS [23] as sources of sensor data. There has been significant work on audio analysis and signal processing. The basic problem of sound classification has been as an active area of research [35]...

Citation Context

...ta was being collected. We then built predictive models for activity recognition using three classification algorithms. The topic of accelerometer-based activity recognition is not new. Bao & Intille =-=[3]-=- developed an activity recognition system to identify twenty activities using bi-axial accelerometers placed in five locations on the user’s body. Additional studies have similarly focused on how one ...

Citation Context

...terms of the accuracy of classification [7, 8, 9]. For example, more than 20 human activities including walking, watching TV, running, stretching and so on can be recognized with fairly high accuracy =-=[10]-=-. While it is still not clear what kind of classification algorithms works the best, most of the existing works require accelerometer sensor(s) to be installed on pre-identified position(s) near human...

Citation Context

...form for wearable context-aware computing. Previous feature extraction studies examining accelerometer data have shown that it is a viable input for detecting user states when it is worn on the wrist =-=[1]-=-. Motivated by other possible sensor platform locations, especially with mobile communication devices such as a cell phone or PDA, we designed a study to investigate the dependency of the eWatch class...