Interview with Dr. Spincer Koh, CEO, Gorilla Technology Group
Dr. Spincer Koh has more than 20 years of experience in the high-tech industry. He also works as a consultant for international broadcast organizations, governments in Asia and the Middle East, and American media groups. He has been appointed the Chairman of China Broadband Application and Promotion Association and involved with New Digital Media Business Development on DVB-H, 3G-TV, WiMAX-TV and Interactive News Media.
Gorilla Technology, a privately held company established in 2000, specializes in video intelligence and IoT technology, supporting a wide range of video-centric and content management applications for retail, enterprise and surveillance. Their machine learning and deep learning video analysis algorithms identify, analyze and extract information from digital content to drive business intelligence solutions and automation.
You started Gorilla Technology back in the year 2000. Tell us about how you have seen the video intelligence and IoT technologies evolve?
Gorilla Video IoT Platform Evolution
Media digitalization and frame-based content analytics
- Gorilla offered a complete premium media management solution including ingestion, quality measurement, video analysis, frame-based meta-tagging, 3-tiered archiving, media search and access, video retrieval, repurposing, and sharing.
- SD formatting within digital content enabled Gorilla to implement shot-and-scene detection technology to retrieve key information from video frames. Examples of this are celebrities being cast in movies which become more efficient from storage and data archiving locations.
Media content to surveillance video analytics
- Gorilla’s managed media content continued to grow and became the leading media and broadcasting archiving platform company.
- As 2K, 4K, and 8K video formats were released, Gorilla was able to generate better metadata with increased bit and frame rates per second on all media.
- The emergence of digital security cameras offered the possibility of better video analytical results. Strong market demand allowed Gorilla to develop video analytics utilizing various machine learning networks for security camera footage.
Deep learning algorithms accelerates video analytic development
- HD format adoption within the CCTV industry enables Gorilla to offer better accuracy with their video analytic solutions.
- Gorilla is proactively developing its video analytic deep learning algorithmic technology.
- Gorilla leverages its IoT sensor and device technology to retrieve invisible data from video streams and provide a comprehensive solution for different verticals.
- Gorilla develops Edge-based AI framework and Hybrid Edge-Cloud platforms to minimize network requirements while maximizing the outcome of actionable insights from surveillance video.
- Gorilla Deep Learning Optimization Technology on pure x86 graphic processing frameworks accelerate computing capability by more than 50%, thus improving overall video analytics across more channels simultaneously and capturing more frames per second, resulting in much better cost performance for customers.
How does Gorilla Tech’s video intelligence and IoT technology allow brands to interact with their customers through their buyer journey?
Video analytics identify a customer’s demographic. RFID tags on merchandise and store beacons log the customer’s interaction with specific products and measure their shopping routes. Using the customer’s shopping records and in-store campaigns, targeted advertising is tailor-made to increase a customer’s propensity to purchase.
Could you give us an example of how Gorilla Tech’s offering can help simplify the decision-making process in retail?
MAP, the leading lifestyle retailer with over 2,200 retail stores in Indonesia, uses Gorilla Smart Retail to analyze customer demographics, traffic, shopping route, and popular product purchases.
The integration with POS data further provides real-time customers shopping behavior versus merchandise and brand information in detail. Once the data is collected from all end-points at each Planet Sport chain store, one of MAP’s retail sub-brands, the store layout is altered to expand a popular brand counter area and several brand locations are re-arranged to achieve better sales results. It also helps to reach the just-in-time merchandise inventory control to minimize costs.
How can brands leverage Gorilla Tech’s IoT platform to align in-store marketing with their digital strategy?
Real-time data of shoppers interacting with products and shopping routes extracted from videos provide more insights into customer trends, which when combined with e-commerce statistics and POS data allow store managers to run efficiently and targeted in-store marketing campaigns.
How can VeMo’s technology be molded to suit different businesses and their unique needs?
Gorilla’s VeMo Matrix, available aims to transform data into actionable, real-time insights of people, vehicle and object identification & verification capabilities and location-based analytics to create value and drive new opportunities for businesses.
- Airports: Manage and monitor passengers, employees, vendors and vehicles across airports monitoring watchlist and suspicious activity crossing restricted areas and sending alert notifications.
- Commercial Buildings: Manage building entrances using facial recognition and monitor all areas for suspicious activity.
- Correctional Agencies: Monitor inmates, visitors, and vehicles and alert administrators of abnormal behavior and intrusions.
- Constructions sites: Manage site worker and vehicle access to prevent intruders entering and avoid accidents.
- Transportation: Monitor vehicle for traffic control and track vehicle routes city-wide.
- Train/MRT Stations: Monitor passenger traffic and suspicious activity allowing for immediate response alerts.
What kind of safety measures does your company ensure while storing enterprise data?
We apply an added layer of network security to detect and locate APT and malware attacks through data access flows which cover login authentication, data encryption on top of firewall and VPN protection.
Take us through VeMo’s work in vehicle identification in real-time.
Smart Patrol, cloud-based license plate recognition solution, is installed and integrated in patrol vehicles. The system allows law enforcement agencies to position camera setups to target and scan license plate information from oncoming, side and rear-facing camera traffic. Smart Patrol improves patrol duty by replacing manual detection operations with real-time streaming, parking violation video analytics, and alert system notifications to deter crime and assist with investigations. Anti-pollution, and parking management measurements have also been implemented as part of Smart City projects. Vehicle detection and traffic control classification are widely used as well.
Gorilla’s Fast Video Solution (FVS) identifies vehicle color and models which helps authorities to locate targeted vehicles through the national transportation surveillance network more efficiently, as well as logistic fleet control.
How does Gorilla Tech’s video analytics work for surveillance? How does the video technology define and identify ‘suspicious’ activity?
Gorilla has developed different video analytic algorithms for surveillance purposes across different verticals including people, behavior, vehicle, object and business intelligent analytics. For example, in correctional facilities, behavior analytics like intrusion, loitering, tripwire, crowd gathering, fights, directional movements and camera tampering has been used to identify suspicious activity in prisons.
What is the company policy on privacy and how does it fit into Gorilla Technology’s work?
Gorilla only keeps encrypted unreadable data from video analytics to prevent data leaks. All original source data and video files are stored at the client’s location.
Tell us about some of your key partner companies and how you integrate technology
Gorilla currently works with key technology and channel partners to integrate and deliver solutions globally.
- Technology partners:
- Channel Partners:
- System Integrator: One SI partner for one vertical in each region. For example, AFN in Australian retail, TSM for Australian security, SMI in Indonesian retail, Ricoh in Thailand retail, FGC in Saudi Arabian security, Foxxtel in Egyptian security, etc.
Thank you, Spincer! That was fun and hope to see you back on AiThority soon.