Deep Learning in Computer Vision, Machine Learning

Digica is an independent Data Science and Artificial Intelligence company. We help our clients by applying the latest Data Science and AI tools and techniques to their systems and products. This leads to highly customized solutions for our customers. Based on our own extensive research program our experienced team makes sure that the best available approach is applied in any situation.

We have deep expertise in the field of image processing including Deep Learning for Computer Vision and leading-edge commercial implementation of Synthetic Imaging. Our work also covers the fields of Financial Trading, Audio Analysis and Predictive Maintenance.

As a group of engineers and engineering managers we are experts in the practical application of Artificial Intelligence and Machine Learning. We help to innovate by applying Machine and Deep Learning techniques. If you believe your product or solution can be improved by AI, but to do so seems to be too complex or risky, our well-experienced team will help you to overcome challenges and will guide you into the World of “new electricity”.

 

 
$10,000+
 
$50 - $99 / hr
 
10 - 49
 Founded
2017
Show all +
Lodz, Poland
headquarters
  • Digica Poland, ul. Brukowa 12
    Lodz, LO 91-341
    Poland
other locations
  • Atlantic Business Center, Atlantic St.
    Broadheath, Altrincham, GTM WA14 5NQ
    United Kingdom
  • 1201 3rd Ave. Ste. 3400
    Seattle, WA 98101
    United States

Portfolio

Key clients: 
LEGO, T-Mobile US, TomTom, Deloitte, Fujitsu
Target classification with FMCW radar signal Image

Target classification with FMCW radar signal

Enigma Pattern developed a method for training CNN models to classify objects in FMCW radar data. The team developed a unique way of processing domain-specific signals. The model manages to self classify versatile objects.

TECHNOLOGIES

Convolution Neural Networks, TensorFlow, scikit-learn, scikit-image

RESULTS

Scalar data (azimuth, distance, speed, etc.), 1s

observation:
gain in Recall: 5%-10%

Scalar data + microDoppler spectra, 1s observation:
gain in Recall: 20%-25%

Scalar data + microDoppler spectra, 5s observation:
gain in Recall: 30%
Preventive maintenance – No Trouble Found modeling Image

Preventive maintenance – No Trouble Found modeling

Using a continuous stream of information, individual to every mobile device, consisting of:

- internal state (OS, make, model, set of installed applications)

- user's behavior (number and time of phone calls, number and time of SMSs,  WiFi on / off, etc.)

 

Enigma Pattern trained a

model that predicts if a specific mobile device will crash in the near future.

Additionally the model was able to advise customer service on preventative maintenance actions such as upgrade/downgrade of OS, removal of applications etc.

 

Results: 

Model predicting failure of a mobile device with 91% accuracy. 



Technologies: SHAP, decision trees (XGBoost)

Deep Learning in detection and classification of any LEGO elements Image

Deep Learning in detection and classification of any LEGO elements

Project description:

Enigma Pattern has created a new type of Deep Neural Network that allows real time detection and classification of LEGO items out of 400.000 different LEGO elements. For training it requires only synthetic images.

 

Challenges:

 

Lack of real pictures of LEGO elements taken in different lighting conditions reflecting possible scenarios of children’s play. 

  • Speed and accuracy of the selected method on mobile devices.

  • Technologies and methods used:

    TensorFlow, TensorFlow Lite, CoreML, Unity 3D, our proprietary “Synthesis” method 


    Results: mAP 89%, real time detection and classification experience

    Deep Learning in classification of electronics items Image

    Deep Learning in classification of electronics items

    Project description:

    One of the biggest distributors of Electronics in the World has employed Enigma Pattern to create a solution that will allow detection and classification of all 300,000 products using deep neural networks.

    Challenges:
    The main challenge was the number of pictures available - 15 per SKU, the secondary challenge was related to the very high quality of those pictures

    as they were taken in a professional environment in perfect lighting conditions.

    Technologies and methods used:

    Neural Networks, Keras, TensorFlow, Triplet Loss, VGG16, Computer Vision techniques

     

    Results:

    Classification of main categories: 97%

    Classification of subcategories: 92%

     

    Classification of a SKU: 87%

     

    Machine Learning in detecting chemical compounds in measurements of Ion Mobility Spectrometry Image

    Machine Learning in detecting chemical compounds in measurements of Ion Mobility Spectrometry

    Overview: Enigma Pattern developed CNN models capable of detecting predetermined set of chemical compounds in IMS spectra of gaseous samples. The method is resilient to noise and changes in measurements originating from different ambient conditions. 

     

    The secondary goal was to determine which parts of the measured

    spectrum is the most important to classification process. EP’s approach to dataset augmentation allowed for
    efficient learning on limited training dataset, whose collection and labeling is especially expensive,
    compared to other domains.

     

    Results: Enigma Pattern has identified a subset of features comprising 25% of the original information. Model trained on such data retained 

    over 96% of full model’s detection performance.

    Technologies: Tensorflow, Keras

    Deep Learning in Road Signs Detection Image

    Deep Learning in Road Signs Detection

    Road Signs Detection - Mapping systems for autonomous vehicles:

    Road mapping cars receive a vast quantity of visual data per second. It is crucial to analyze such data efficiently and to process it quickly.

    The model enabled immediate visual recognition and segmentation of road-related signs and markings through neural network systems. Such processes proved to be essential e.g. for

    autonomous cars’ driving systems, as they required precise road markings for correct mode of operation.

     

    Results:

     

    The system achieved 90% accuracy in visual signs’ recognition with set Jaccard Index parameters preserved.

     

    Technologies:

    Python, Keras, TensofFlow on GPU.

     

    SSD Algorithms:  Single Shot MultiBox Detector

    Deep Learning in Image Reconstruction Image

    Deep Learning in Image Reconstruction

     

    Fixed Pattern Noise Removal from Thermal Images:
    The objective of the project was to remove the fixed pattern noise on thermal cameras which are susceptible to both external (environmental) and internal (built-in) conditions.  

    Challenges:
    The primary concern was to remove the noise from the screen while preserving the actual image, this means that no additional data (hallucinations)

    should appear after noise removal. Additionally, since there are three different types of noises affecting thermal images, each of the noises had to be removed separately. 

    Results:

    The low-frequency noise was decreased by 80%.
    The number of artifacts was decreased by 30%.
    The high-frequency noise was reduced by 20%.

    Target of achieving threshold of 25 dB achieved, where

     

    30 dB is believed as fully denoised picture.

     

    Technologies:

     

    Keras, TensorFlow

    Deep Learning – Synthetic Images in image recognition

    Development of a process to train CNN with synthetic images
    Enigma Pattern was contracted to develop an unique method to generate and train convolutional neural network models basing on synthetic images. The entire process comprised:
    - development and preparation of the environment basing on Caffe
    - development of steering scripts for simulation of a natural environment using SynCity
    - development

    of variants of the image object classifiers
    - modification of hyperparameters of the network in order to improve detection precision
    - improvement of results by transformation of synthetic images  

    Results:
    Repeatable process of building neural networks
    based on synthetic images
    See a separate presentation about results of this project.

    Technologies:
    Caffe, Unity3D

    Reviews

    Sort by

    Machine Learning Workshop for ICT Co

    "It was a very collaborative and positive experience working with Enigma Pattern."

    Quality: 
    5.0
    Schedule: 
    5.0
    Cost: 
    5.0
    Willing to refer: 
    5.0
    The Project
     
    Less than $10,000
     
    Jan. 2020 - Jan. 2020
    Project summary: 

    Enigma Pattern hosted a machine learning workshop for a global ICT company. The program was customized for the engineers and the company that attended.

    The Reviewer
     
    10,000+ Employees
     
    Lodz, Poland
    Principal Software Developer, ICT Company
     
    Verified
    The Review
    Feedback summary: 

    The agenda met project requirements and the training was successful. They plan to have another event soon. Enigma Pattern was easy to communicate with over phone and email. 

    The client submitted this review online.

    BACKGROUND

    Please describe your company and your position there.

    Principle Software Engineer and Team Leader for the leading Japanese global information and communication technology (ICT) company, offering a full range of technology products, solutions and services.

    OPPORTUNITY / CHALLENGE

    For what projects/services did your company hire Enigma Pattern?

    Enigma Pattern did a ML workshops for a team of 8 engineers.

    What were your goals for this project?

    To gain basic ML knowledge.

    SOLUTION

    How did you select this vendor?

    We compared the offers from different vendors and chose Enigma Pattern based on that.

    Describe the project in detail.

    Enigma Pattern did custom ML workshops for us.

    What was the team composition?

    3 people. Head of Data Science in Enigma Pattern and two Data Scientists.

    RESULTS & FEEDBACK

    Can you share any outcomes from the project that demonstrate progress or success?

    We were satisfied with the workshops and consider to have a stage 2 training for the same team.

    How effective was the workflow between your team and theirs?

    The communication with the vendor was easy. We exchanged few emails and had few phone calls.

    What did you find most impressive about this company?

    It was a very collaborative and positive experience working with Enigma Pattern. They prepared the agenda for the training according to our requirements.

    Are there any areas for improvement?

    We would like to receive materials after the training in the future.

    5.0
    Overall Score
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 5.0 Cost
      Value / within estimates
    • 5.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer

    Machine Learning Model Development for IoT Platform

    "It is very clear that their work has shown us and our clients sufficient value."

    Quality: 
    4.0
    Schedule: 
    5.0
    Cost: 
    4.0
    Willing to refer: 
    5.0
    The Project
     
    $10,000 to $49,999
     
    Sept. - Dec. 2019
    Project summary: 

    Enigma Pattern provided Artificial Intelligence Markup Language (AIML) and data science expertise to an IoT services company. Their AIML models then led to the creation of crucial reports on data trends.

    The Reviewer
     
    51-200 Employees
     
    Milpitas, California
    CEO, IoT Platform
     
    Verified
    The Review
    Feedback summary: 

    Internal subject matter experts relied on the AIML models to reveal hidden trends, patterns, and information for an end client. Flexible and communicative, Enigma Pattern’s high-caliber team is now a regular partner for assisting with AIML and data science work.

    A Clutch analyst personally interviewed this client over the phone. Below is an edited transcript.

    BACKGROUND

    Introduce your business and what you do there.

    I’m the CEO of an IoT software company providing end-to-end services for global customers.

    OPPORTUNITY / CHALLENGE

    What challenge were you trying to address with Enigma Pattern?

    We were trying to take our data analysis to the next level for one of our larger clients. Our team was gathering a lot of data to make sudden reporting and value derivations for them. We knew there was more we could unpack with their data, so we hired Enigma Pattern to help us meet this challenge.

    SOLUTION

    What was the scope of their involvement?

    They provided hardcore Artificial Intelligence Markup Language (AIML) and data science expertise. Looking at the data, they would make decisions on what the data told us using different AIML models and techniques for pattern-seeking.

    Our subject matter experts would work with them and request data to try to modify these models, looking to produce more relevant information for our customer’s business. Using this, we could then derive a set of very valuable reports for our client.

    How did you come to work with Enigma Pattern?

    I had an existing relationship with Rafal (CEO, Enigma Pattern). We knew each other in a previous life, and I knew he was very easy to work with and would find very high-caliber individuals for me to work with.

    How much have you invested with them?

    We invested somewhere in the £30,000–£50,000 range (approximately $39,000–$65,000 USD).

    What is the status of this engagement?

    The project lasted from September–December 2019, but we’re looking to start further projects with them.

    RESULTS & FEEDBACK

    What evidence can you share that demonstrates the impact of the engagement?

    They provided us a set of valuable reports, graphs, and charts that described the trends, patterns, and information in our data. This information was all gleaned from the modeling they had done. It is very clear that their work has shown us and our clients sufficient value, so much so that we’re moving forward to a second and third phase of this project.

    How did Enigma Pattern perform from a project management standpoint?

    They performed well. I was slightly removed from the project, but the overall feedback I received was positive. 

    What did you find most impressive about them?

    They were very flexible. They were ready to work with us regardless of the amount of work we had, and they were easy to work with in a way that many partners, unfortunately, are not.

    Are there any areas they could improve?

    No, not in any way I know. I was a step removed, but no problems were ever escalated to me. We’re very happy with Enigma Pattern.

    5.0
    Overall Score
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 4.0 Cost
      Value / within estimates
    • 4.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer

    AI Services for Thermal Imaging Systems Company

    "There aren’t that many people who have as good a knowledge base as the Enigma Pattern team."

    Quality: 
    4.0
    Schedule: 
    3.0
    Cost: 
    4.0
    Willing to refer: 
    4.0
    The Project
     
    Confidential
     
    June 2018 - Apr. 2019
    Project summary: 

    Enigma Pattern provided data science services to transition analytics for a thermal imaging systems company. They researched and developed neural networks system to integrate AI into thermal imaging systems.

    The Reviewer
     
    1001-10,000 Employees
     
    Santa Barbara, California
    Technical Lead, Thermal Imaging Company
     
    Verified
    The Review
    Feedback summary: 

    Enigma Pattern delivered valuable services that augmented system operations. Despite initial challenges, the team supported advancements to the collaboration with improved project management skills. They provide a valuable depth of knowledge and analytical skills to enable continued partnership.

    A Clutch analyst personally interviewed this client over the phone. Below is an edited transcript.

    BACKGROUND

    Introduce your business and what you do there.

    I’m the technical lead of an international company that builds thermal imaging systems.

    OPPORTUNITY / CHALLENGE

    What challenge were you trying to address with Enigma Pattern?

    We wanted their help in applying AI to thermal imaging systems.

    SOLUTION

    What was the scope of their involvement?

    Enigma Pattern provided data science services and neural networks system development. We were active in this project for six months. I can’t say anything in more detail than that. We’ve done more than one project with Enigma Pattern and are working on another one right now.

    How did you come to work with Enigma Pattern?

    My boss, who is the CTO, ran into Enigma Pattern at a trade show. There was interest in following up, so we had initial meetings with their team.

    What is the status of this engagement?

    Our work began a year and a couple of months ago. We’ve done more than one project with them and have a good working relationship with Enigma Pattern.

    RESULTS & FEEDBACK

    What evidence can you share that demonstrates the impact of the engagement? 

    This project was to replace conventional AI analytics with AI-based analytics. They were able to perform at about 30%, so there was a significant improvement in performance.

    How did Enigma Pattern perform from a project management standpoint?

    They provided a weekly report that we reviewed section by section. There were rocky moments at times but it was good in the end; they’re in Poland and we’re in the US but their project management was adequate.

    What did you find most impressive about them?

    There aren’t that many people who have as good a knowledge base as the Enigma Pattern team. They have several Ph.D. holders who are quite knowledgeable on their team. Their strength was their knowledge of neural networks, analytical and data science capabilities.

    Are there any areas they could improve?

    They could improve their project management skills. We didn’t encounter technical issues, rather issues with project management. They definitely made strides to improve it later on but there were definitely some growing pains.

    4.0
    Overall Score
    • 3.0 Scheduling
      ON TIME / DEADLINES
    • 4.0 Cost
      Value / within estimates
    • 4.0 Quality
      Service & deliverables
    • 4.0 NPS
      Willing to refer

    AI Development Services for Toy Company

    “Enigma Pattern has delivered on time, within budget, and to our expectations.”

    Quality: 
    4.0
    Schedule: 
    4.5
    Cost: 
    4.0
    Willing to refer: 
    4.0
    The Project
     
    Confidential
     
    Jan. 2018 - Ongoing
    Project summary: 

    Enigma Pattern provides development services to a toy company. They’re helping to develop an AI product.

    The Reviewer
     
    10,000+ Employees
     
    Denmark
    Lead Producer, Consumer Goods Company
     
    Verified
    The Review
    Feedback summary: 

    Their well-qualified team has a unique approach to resolving AI tasks and internal problems. Enigma Pattern provides actionable insights, adheres to deadlines, and regularly follows up. 

    A Clutch analyst personally interviewed this client over the phone. Below is an edited transcript.

    BACKGROUND

    Introduce your business and what you do there.

    I work for a toy company, and I’m responsible for sourcing external technologies to use in our products.

    OPPORTUNITY / CHALLENGE

    What challenge were you trying to address with Enigma Pattern?

    We had some AI tasks to solve.

    SOLUTION

    What was the scope of their involvement?

    Enigma Pattern provides development for an AI product. I cannot disclose any details.

    How did you come to work with Enigma Pattern?

    We chose them because they had good insights into our problem. They also had some qualified people to solve that problem for us, and they made a good pitch.

    What is the status of this engagement?

    We started working with Enigma Pattern at the beginning of 2018. The project is ongoing.

    RESULTS & FEEDBACK

    What evidence can you share that demonstrates the impact of the engagement?

    Enigma Pattern has delivered on time, within budget, and to our expectations.

    How did Enigma Pattern perform from a project management standpoint?

    They’re quite good at following up, and they’ve met the main and interim deadlines. They provide us with some Excel sheets for project management. We use tools like Asana and Jira internally, but not with Enigma Pattern.

    What did you find most impressive about them?

    They’re on par with other service providers we’ve worked with in the same field. They have a unique approach to solving our problem, which is quite good.

    Are there any areas they could improve?

    No, there’s nothing off the top of my mind.

    4.5
    Overall Score
    • 4.5 Scheduling
      ON TIME / DEADLINES
    • 4.0 Cost
      Value / within estimates
    • 4.0 Quality
      Service & deliverables
    • 4.0 NPS
      Willing to refer

    AR/VR Dev for Telecommunications Company

    “They have great ethics, which made a big difference for us. We’ll definitely consider them for our next project.”

    Quality: 
    5.0
    Schedule: 
    5.0
    Cost: 
    5.0
    Willing to refer: 
    5.0
    The Project
     
    Confidential
     
    Apr. - Aug. 2019
    Project summary: 

    Enigma Pattern is developing an AR/VR tool for a telecommunications company. They’ve created an adaptable and scalable solution using AI for detection and data manipulation.

    The Reviewer
     
    10,000+ Employees
     
    Bellevue, Washington
    Director of Product Development, Telecommunications Company
     
    Verified
    The Review
    Feedback summary: 

    Though the product hasn’t launched yet, the internal team expects it to have a big impact on existing products. Enigma Pattern delivered quality work and remained professional throughout the project. The team communicated regularly and delivered everything on time.

    A Clutch analyst personally interviewed this client over the phone. Below is an edited transcript.

    BACKGROUND

    Introduce your business and what you do there.

    I’m the director of product development at a telecommunications company.

    OPPORTUNITY / CHALLENGE

    What challenge were you trying to address with Enigma Pattern?

    We were looking for a partner with excellent AI skills to help us develop a new product for our team and marketing organization

    SOLUTION

    What was the scope of their involvement?

    They’re helping us develop an AR/VR product. We described the product and high-level specifications of what we wanted to achieve. They dug deeper and developed more technical details to share with our team. After some back and forth, they created the wireframe. They worked on everything, from A to Z. After determining the technical requirement and documentation, they began coding. It’s an out-of-the-box experience in AI detection and data manipulation on behalf of the customer.

    Although our most recent product hasn’t launched yet, and I can’t disclose much about it, we’ve worked on other projects with Enigma Pattern. One is a broadband solution we launched in rural areas. The customer can use an AV app to determine if one of our products is placed properly.  

    What is the team composition?

    I’ve worked with Rafael and the technical lead. They were the focal point and proxy between us and the developers.

    How did you come to work with Enigma Pattern?

    We had several teams submit an RFP. Enigma Pattern came out on top for several reasons: they were cheaper, good-quality products, and had significant previous experience. Above all else, the quality of their work led to us choosing them over others.

    What is the status of this engagement?

    The project lasted from April—August 2019.

    RESULTS & FEEDBACK

    What evidence can you share that demonstrates the impact of the engagement? 

    We definitely believe that this new development will uplift our products. At this point, I don’t have any metrics, but internally, we’re pleased and think it will make a difference. The quality of Enigma Pattern’s code is great. We’ve run it by other teams and have determined that it’s adaptable, flexible, and most importantly, scalable.

    How did Enigma Pattern perform from a project management standpoint?

    They’re very professional and have met all of our deadlines. If there’s any delay, they would tell me way in advance. They’re always on top of it. Their process is really rigorous in terms of updating and communicating with the customer.

    What did you find most impressive about them?

    They have great ethics, which made a big difference for us. We’ll definitely consider them for our next project.

    Are there any areas they could improve?

    Not really, nothing seemed to have a negative impact.

    Do you have any advice for potential customers?

    Be specific on the scope. From learning with them, we learned to document everything and to be formal. Don’t boil the ocean, because you don’t want to spend more money or time on the project than you have to.

    5.0
    Overall Score
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 5.0 Cost
      Value / within estimates
    • 5.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer

    Software Dev for Radar Manufacturer

    "Their management was well organized and responsive."

    Quality: 
    5.0
    Schedule: 
    5.0
    Cost: 
    5.0
    Willing to refer: 
    5.0
    The Project
     
    Confidential
     
    Mar. 2019 - Jan. 2020
    Project summary: 

    Enigma Pattern was tasked with investigating convolutional neural networks (CNN) and implementing CNN algorithms into the product code of a radar manufacturer. 

    The Reviewer
     
    501-1,000 Employees
     
    Montreal, Canada
    Engineering Manager, Radar Manufacturer
     
    Verified
    The Review
    Feedback summary: 

    While the scope and results of the project remain confidential, the client commends the team for their well-organized project management and an impressive response rate. Enigma Pattern's dedication to pleasing the client and flexibility in the face of changes both make them a remarkable partner.

    The client submitted this review online.

    BACKGROUND

    Please describe your company and your position there.

    We manufacture radars and I was the project supervisor and customer

    OPPORTUNITY / CHALLENGE

    For what projects/services did your company hire Enigma Pattern?

    We hired Enigma Pattern to help us investigate the use of convolutional neural networks (CNN) and deep-learning for radar

    What were your goals for this project?

    Develop CNN algorithms for radar

    SOLUTION

    How did you select this vendor?

    Previously used by us

    Describe the project and the services they provided in detail.

    Confidential

    RESULTS & FEEDBACK

    Can you share any information that demonstrates the impact that this project has had on your business?

    Confidential

    How was project management arranged and how effective was it?

    Their management was well organized and responsive.

    What did you find most impressive about this company?

    Eagerness to please the customer and flexibility/adaptability to changes

    Are there any areas for improvement?

    No

    5.0
    Overall Score
    • 5.0 Scheduling
      ON TIME / DEADLINES
    • 5.0 Cost
      Value / within estimates
    • 5.0 Quality
      Service & deliverables
    • 5.0 NPS
      Willing to refer