This guide outlines the steps for getting started with an evaluation of Trueface's software. Assuming you have already been issued an evaluation token (if not, please contact the team), identify the right solution for your use case, verify your environment meets the minimum hardware requirements, and get started with the appropriate solution guides.
Step 1: Determine which solution is the most appropriate for your use case and deployment.
We currently offer three solutions, each tailored towards a different use case or deployment environment. Below is a quick summary of each solution. If you're still unsure which solution is best for your use case, contact sales or read through the main documentation page which has more information and download links for all solutions.
Pangiam Trueface SDK
The Trueface SDK is a building block at the core of all our products and offers the highest level of performance and most freedom by supporting various configurations and options. The SDK is shipped as a static or dynamic library and offers a C++ API and Python bindings. It supports a variety of targets and architectures, including x86_64 and ARM, as well as CPU and GPU deployments. As such, it can be deployed anywhere, from large cloud CPU / GPU clusters, to edge devices such as iPads and Android phones, and everything in between. The Trueface SDK provides the most flexibility to build your application the way you need as all API functions and modules are accessible (with very fine granularity), allowing you to build custom pipelines for your use case. We advise using the SDK where performance is critical (latency and throughput) or when processing large amounts of data. Additionally, the SDK should be the first choice for edge deployments.
Trueface SDK Getting Started Guide
Pangiam Trueface On-Prem (PTOP)
The Trueface On-Prem is a wrapper around the Trueface SDK which packages the SDK in a docker container and exposes the SDK face recognition functionality through a REST API server. The PTOP adds a layer of abstraction above the SDK such that a single convenient REST API endpoint can invoke an entire pipeline of SDK methods to achieve complex tasks such as face recognition, without the caller needing to worry about the internal workings. Although this solution can be deployed on any platform supporting docker, including on-premise deployments, it is ideal for cloud deployments such as in an AWS environment. This solution is useful for when it is necessary to transfer data over a network as it provides a convenient REST API server. Ex: A web browser application accessed on mobile phones sends images via REST from the onboard camera to a server running an On-Prem instance to run face recognition queries against a database.
Trueface On-Prem Getting Started Guide
Pangiam Trueface Cloud
The Trueface Cloud is a Software as a Service (SaaS) offering that provides the same functionality as PTOP but deployed to the cloud. All the operations are available via the REST API on a custom domain. It is designed as a turn-key solution in which Pangiam will managed all the resource provision and scaling to meet your performance needs. The Trueface Cloud is designed to provide a fast and reliable performance regardless of dataset size, from a few millions to over one billion facial vectors. For this offering, step 2 can be skipped.
Step 2: Ensure your target hardware meets the minimum and recommended hardware requirements.
Visit the following page and ensure your runtime hardware meets the minimum required hardware specified for the selected solution.
Minimum & Recommended Hardware Requirements