We are pleased to announce that the HASCA (Human Activity Sensing Corpus and Applications) Workshop will take place as part Ubicomp2021 in September 2021.
HASCA is one of the largest workshops in Ubicomp, it has been held for 9 years.

HASCA WORKSHOP IN UBICOMP2021


Date TBD (September 25 or 25, 2021, virtual event)


The objective of this workshop is to share the experiences among
researchers about current challenges of real-world activity
recognition with newly developed datasets and tools, breaking through
towards open-ended contextual intelligence.

This workshop discusses the challenges of designing reproducible
experimental setups, the large-scale dataset collection campaigns, the
activity and context recognition methods that are robust and adaptive,
and evaluation systems in the real world.

As a special topic of this year we will reflect on the challenges to
recognize situations, events and/or activities among the statically
predefined pools and beyond - which is the current state of the art -
and instead we will adopt an "open-ended view" on activity and context
awareness. This may result in combinations of the automatic discovery
of relevant patterns in sensor data, the experience sampling and
wearable technologies to unobtrusively discover the semantic meaning
of such patterns, the crowd-sourcing of dataset acquisition and
annotation, and new "open-ended" human activity modeling techniques.

We expect the following domains to be relevant contributions to this
workshop (but not limited to):

- *Data collection*, *Corpus construction*.
Experiences or reports from data collection and/or corpus construction
projects, including papers which describes the formats, styles and/or
methodologies for data collection. Cloud-sourcing data collection and
participatory sensing also could be included in this topic.

- *Effectiveness of Data*, *Data Centric Research*.
There is a field of research based on the collected corpora, which is
so called "data centric research". Also, we call for the experience of
using large-scale human activity sensing corpora. Using large-scale
corpora with an analysis by machine learning, there will be a large
space for improving the performance of recognition results.

- *Tools and Algorithms for Activity Recognition*.
If we have appropriate tools for the management of sensor data,
activity recognition researchers could have more focused on their
actual research theme. This is because the developed tools and
algorithms are often not shared among the research community. In this
workshop, we solicit reports on developed tools and algorithms for
forwarding to the community.

- *Real World Application and Experiences*.
Activity recognition "in the lab" usually works well. However, it does
not scale well with real world data. In this workshop, we also solicit
the experiences from real world applications. There is a huge gap
between "lab" and "real world” environments . Large-scale human
activity sensing corpora will help to overcome this gap.

- *Sensing Devices and Systems*
Data collection is not only performed by the "off-the-shelf" sensors
but also the newly developed sensors which supply information which
has not been investigated. There is also a research area about the
development of new platform for data collection or the evaluation
tools for collected data.

In light of this year's special emphasis on open-ended contextual
awareness, we wish cover these topics as well:

- *Mobile Experience Sampling*, *Experience Sampling Strategies*.
Advances in experience sampling approaches, for instance intelligent
user query or those using novel devices (e.g. smartwatches), are
likely to play an important role to provide user-contributed
annotations of their own activities.

- *Unsupervised Pattern Discovery*.
Discovering meaningful patterns in sensor data in an unsupervised
manner can be needed in the context of informing other elements of the
system by inquiring the user and by triggering the annotation with
crowd-sourcing.

- *Dataset Acquisition and Annotation*, *Crowd-Sourcing*, *Web-Mining*.
A wide abundance of sensor data is potentially within the reach of
users instrumented with their mobile phones and other
wearables. Capitalizing on crowd-sourcing to create larger datasets in
a cost effective manner may be critical to open-ended activity
recognition. Many online datasets are also available and could be used
to bootstrap recognition models.

- *Transfer Learning*, *Semi-Supervised Learning*, *Lifelog Learning*.
The ability to translate recognition models across modalities or to
use minimal forms of supervision would allow to reuse datasets in a
wider range of domains and reduce the costs of acquiring annotations.

- *Deep Learning*
Together with the big success of deep learning in other AI domain, deep
learning models are gradually playing an important role in activity
recognition as well.
AREAS OF INTEREST

Human Activity Sensing Corpus
Large Scale Data Collection
Data Validation
Data Tagging / Labeling
Efficient Data Collection
Data Mining from Corpus
Automatic Segmentation
Performance Evaluation
Man-machine Interaction
Noise Robustness
Non Supervised Machine Learning
Sensor Data Fusion
Tools for Human Activity Corpus/Sensing
Participatory Sensing
Feature Extraction and Selection
Context Awareness
Pedestrian Navigation
Social Activities Analysis/Detection
Compressive Sensing
Sensing Devices
Lifelog Systems
Route Recognition/Detection
Wearable Application
Gait Analysis
Health-care Monitoring/Recommendation
Daily-life Worker Support
Deep Learning