Synetiq has a large dataset of videos and triggered human responses, even emotions. We are always improving our products with exciting features, and we would appreciate a motivated new colleague to explore some new opportunities.

We expect a highly autonomous candidate that can hack things on her own. Besides, supervision and guidance will be provided on daily basis by our skilled tech and research teams.

Synetiq is a leading neuromarketing media research company measuring people’s emotional reactions to media content. We are a cheerful team of engineers, psychologists, researchers and business experts providing a solution for broadcasters, agencies and advertisers to better understand their audience.

We are currently offering four different internship projects:

1. Ranking/recommending videos

Project description
The proliferation of high speed networks and improvements of data mining techniques have created the possibility for personalized online video advertising. Selecting which video should be presented to each user is still an open challenge.

Scope
Build a ranking/recommender model and training pipeline that applies learning to rank or multi-armed bandit techniques to select the best video for a user or segment. For simplicity, each video is previously encoded as a numerical feature vector.

Skills
Hands on experience with data science techniques in either Python, R or Matlab.
Knowledge of recommender systems or search engines is appreciated.

Seniority
BSc/MSc or demonstrable experience

Start date
Anytime

Length
350 hours (approx 2 months full-time or 4 months part-time)

2. Cluster people based on their response profiles

Project description
Clustering is an unsupervised technique that could help unravel meaningful patterns in data. In this specific use case, the applicant will get responses of a survey from thousands of participants. The survey consists of different sub-questionnaires of variable question types.

Scope
The task is to find clusters of testers that would help identifying groups of testers and potentially shorten the length of the survey in the future. We encourage the applicant to use standard techniques (e.g. k-means clustering). The clustering performance is measured on questionnaire data.

Skills
Hands on experience with data science techniques in either Python, R or Matlab or any analytics platform (Knime, Spotfire, Tableau, Rapidminer).

Seniority
BSc/MSc or demonstrable experience

Start date
Anytime

Length
350 hours (approx 2 months full-time or 4 months part-time)

3. Predict eye tracking

Project description
The use of eye tracking devices makes possible the analysis of human attention on images. These methods are not scalable, hence predicting where humans are looking at based only on the video content has a big potential.

Scope
Build a model and training pipeline that generates the expected heatmap of attention based only on the pixels of the image.

Skills
Hands on experience with deep learning techniques in either Python, R or Matlab.
Hands on experience with computer vision and CNNs.

Seniority
MSc or demonstrable experience

Start date
From 1st of July 2017

Length
350 hours (approx 2 months full time)

4. Aesthetics index implementation by machine learning

Project description
Aesthetics is an important factor in advertising business. We invite the applicant to develop a solution that helps the quantification of aesthetic factors. Neuroscience research tells that people like symmetric faces and faces that do not deviate too much from the population average face.

Scope
The task will be based on this and similar results and will include face recognition and facial feature extraction. We encourage the applicant to identify facial landmarks either by HOG, Haar cascade classifiers and/or deep neural networks.

Skills
Hands on experience with data science techniques in either Python, R or Matlab.
Hands on experience with computer vision and CNNs.

Seniority
MSc or demonstrable experience

Start date
Anytime

Length
350 hours (approx 2 months full-time or 4 months part-time)

Apply to one of these data science internships by specifying which of the 4 you are interested in:

Please click here to apply.