iProov Ltd v. Software Colombia Servicios Informaticos SAS

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: iProov LTD (United Kingdom)
  • Defendant: Software Colombia Servicios Informaticos SAS (Colombian)
  • Plaintiff’s Counsel: Stamoulis & Weinblatt, LLP; Hatch Law PC; Alan Sege Esq., PC
• Case Identification: 1:24-cv-00829, D. Del., 07/17/2024
• Venue Allegations: Plaintiff alleges venue is proper because Defendant committed acts of infringement in the United States by causing U.S.-based computer servers, including those in Ashburn, Virginia, and Columbus, Ohio, to perform elements of the patented methods when using the accused service.
• Core Dispute: Plaintiff alleges that Defendant’s use and implementation of Amazon's Rekognition Face Liveness service infringes two patents related to biometric liveness detection.
• Technical Context: The technology involves authenticating a remote online user by verifying not only their biometric identity but also that they are a live human present at the time of authentication, a critical defense against identity theft via photographs, recordings, or synthetic "deepfakes."
• Key Procedural History: The complaint does not mention any prior litigation, Inter Partes Review (IPR) proceedings, or licensing history related to the patents-in-suit.

Case Timeline

Date	Event
2012-02-21	Earliest Priority Date for '975 and '500 Patents
2015-06-07	U.S. Patent 9,075,975 Issues
2016-10-25	U.S. Patent 9,479,500 Issues
2024-07-17	Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 9,075,975, "Online Pseudonym Verification and Identity Validation," Issued June 7, 2015

The Invention Explained

• Problem Addressed: The patent describes the vulnerability of conventional online authentication methods, including basic facial recognition, to "spoofing attacks" where an unauthorized party uses a photograph or video replay of a legitimate user to gain access (’975 Patent, col. 2:34-38). This circumvents security even if the biometric data itself is correct.
• The Patented Solution: The invention proposes a "liveness" detection method where a server sends control signals to a user's device, causing a light source (such as the device's screen) to modulate its illumination in a specific pattern. A camera on the device records video of the user's face during this process, and the video is sent back to the server for analysis. The server verifies that the changes in light reflected from the user's face correspond in both time and spatial arrangement to the server's control signals, confirming the presence of a live, three-dimensional person in real time (’975 Patent, Abstract; col. 4:16-54).
• Technical Importance: This approach provides a dynamic challenge-response mechanism intended to defeat static spoofing attacks, thereby increasing the security of remote biometric authentication (’975 Patent, col. 2:39-44).

Key Claims at a Glance

• The complaint asserts independent claim 1 and reserves the right to assert others (Compl. ¶20, ¶22-23).
• Exemplary Claim 1 of the '975 Patent recites a method with the following key steps (Compl. ¶19):
  • A server sends control signals to a user device that includes a source of illumination and a camera.
  • The user device receives the signals, modulates its illumination accordingly, and transmits the captured video back to the server.
  • The server receives the video imagery.
  • In a first analyzing step, the server analyzes the video to find frames showing changes in reflected illumination that temporally correspond with the control signals.
  • In a second analyzing step, the server determines the spatial arrangement of these brightness changes and their consistency with what would be expected from a 3D human face.
  • The server uses the first analysis to determine a likelihood that the video was captured in real time.
  • The server uses the second analysis to determine a likelihood that a human face was present.
  • The server generates an authentication response based on both likelihoods.

U.S. Patent No. 9,479,500, "Online Pseudonym Verification and Identity Validation," Issued October 25, 2016

The Invention Explained

• Problem Addressed: Like its parent, the ’500 Patent addresses the need for secure online authentication that can resist spoofing attacks using photographs or video replays (’500 Patent, col. 2:40-50).
• The Patented Solution: The patented method is largely similar to that of the ’975 Patent, involving a server-controlled modulation of a light source and analysis of reflected light from a user’s face. A notable distinction in the claims is the explicit provision that the analytical steps can be performed by "at least one of the server and the user device," affording greater architectural flexibility (’500 Patent, Abstract; col. 8:1-9). The claim also specifies that a "positive authentication response permits the action being sought by the online user to proceed" (’500 Patent, col. 8:6-9).
• Technical Importance: This patent extends the core liveness detection technology by explicitly contemplating system designs where some computational load for the analysis can be handled on the client side, potentially improving performance or enabling different use cases (’500 Patent, col. 8:1-3).

Key Claims at a Glance

• The complaint asserts independent claim 1 and reserves the right to assert others (Compl. ¶29, ¶31-32).
• Exemplary Claim 1 of the '500 Patent recites a method with the following key steps (Compl. ¶28):
  • A server sends control signals to a user device with an illumination source and camera.
  • The user device performs the modulation and transmits video to the server.
  • The server receives the video imagery.
  • Using "at least one of the server and the user device", a first analysis is performed to detect frames with temporally corresponding illumination changes.
  • Using "at least one of the server and the user device", a second analysis is performed to determine the spatial arrangement of brightness changes and consistency with a 3D human face.
  • Using "at least one of the server and the user device", results from the analyses are used to determine first and second likelihoods (real-time capture and human presence).
  • An authentication response is generated, where a positive response permits the user's requested action to proceed.

III. The Accused Instrumentality

• Product Identification: The accused instrumentality is the "Amazon Rekognition Face Liveness" product and service ("Amazon Rekognition") (Compl. ¶12).
• Functionality and Market Context: The complaint alleges that Defendant is a customer and developer that uses and incorporates Amazon Rekognition into its own software and services (Compl. ¶12, ¶22). The alleged infringement occurs when Defendant uses, implements, or offers the service, causing U.S.-based servers to perform the patented methods by connecting to Amazon Rekognition endpoints (Compl. ¶16). The complaint does not provide specific technical details on the operation of Amazon Rekognition beyond alleging that its functionality maps to the elements of the asserted claims.

IV. Analysis of Infringement Allegations

The complaint alleges that Defendant's use of Amazon Rekognition Face Liveness directly infringes at least Claim 1 of the '975 Patent and Claim 1 of the '500 Patent (Compl. ¶20, ¶29). It states that exemplary claim charts are attached as Exhibits C and D, respectively; however, these exhibits were not included with the filed complaint (Compl. ¶20, ¶29).

In prose, the infringement theory for both patents appears to be that the Amazon Rekognition service functions as the claimed "server," which sends control signals to an end-user's device. The device's screen acts as the "source of illumination," which is modulated according to the signals. The device's camera captures video of the user's face, which is then transmitted to and analyzed by the Amazon Rekognition servers (located in the U.S.) to perform the claimed liveness detection analysis (Compl. ¶16, ¶19, ¶28). The allegations for the ’500 Patent specifically note that the analysis may be performed on the server, the user device, or both (Compl. ¶28).

No probative visual evidence provided in complaint.

• Identified Points of Contention:
  • Evidentiary Questions: A central issue will be evidentiary. The complaint makes conclusory allegations that the accused Amazon Rekognition service performs the specific steps of the claims. The plaintiff will need to produce evidence, likely through discovery, detailing the internal technical operations of this "black box" service to prove that it actually performs the claimed multi-step analyses (e.g., analyzing "temporal correspondence" and "spatial arrangement" of reflected light).
  • Scope Questions (Divided Infringement): The alleged infringing method involves actions across multiple entities: Amazon (providing the server), the Defendant (implementing the service), and the end-user (operating the client device). This raises the question of whether direct infringement can be established against the Defendant under the law of divided infringement, which typically requires showing that one party directs or controls the performance of all claim steps.

V. Key Claim Terms for Construction

• The Term: "source of illumination" (’975 Patent, Claim 1; ’500 Patent, Claim 1)

  • Context and Importance: The infringement theory appears to depend on the screen of a user's device (e.g., a smartphone) constituting the "source of illumination." The construction of this term is therefore critical to determining whether the accused system meets this limitation.
  • Intrinsic Evidence for a Broader Interpretation: The specification provides explicit support for a broad definition, stating that sources "include the illuminator or flash associated with a rear-facing camera...and the device display screen itself" (’975 Patent, col. 9:8-12).
  • Intrinsic Evidence for a Narrower Interpretation: A defendant might argue that the term should be limited by the functional requirement that the source be "modulated" in a specific, controlled way. They may point to embodiments describing flashing the screen with "substantially all the screen pixels set to white" (’975 Patent, col. 10:51-54) to argue that a general-purpose screen used for other display functions does not qualify without performing this specific type of modulation.

• The Term: "determining a degree of consistency of the spatial arrangement of the brightness changes...with a spatial arrangement of brightness changes that would be expected from a three-dimensional shape comprising major facial features of a human face" (’975 Patent, Claim 1)

  • Context and Importance: This phrase defines the core 3D liveness check. Practitioners may focus on this term because its construction will determine whether Amazon Rekognition’s proprietary liveness algorithm falls within the scope of the claim, or if it uses a technically distinct method.
  • Intrinsic Evidence for a Broader Interpretation: A plaintiff may argue this term should be construed functionally to cover any algorithm that analyzes reflections from different facial regions to confirm a 3D structure. The specification supports this by describing analysis of "luminance characteristics" and the "spatial distribution of color reflectivity" (’975 Patent, col. 10:38-39; col. 4:44-46).
  • Intrinsic Evidence for a Narrower Interpretation: A defendant could argue the claim requires a more specific analysis than a generic liveness check. They might point to specification language describing how a higher reflective signal is expected from "areas of high reflectivity such as the cornea, spectacles, and the tip of the nose," while an attenuated signal corresponds to "areas of low reflectivity such as the lips or eyebrows," to argue the claim requires this type of specific spatial comparison (’975 Patent, col. 12:37-43).

VI. Other Allegations

• Indirect Infringement: The complaint alleges both induced and contributory infringement.
  • Inducement is premised on allegations that Defendant has "full knowledge," designs its software to use Amazon Rekognition, and provides "on screen visual guidance and online instructional materials" that instruct its customers to use the service in an infringing manner (Compl. ¶22, ¶31).
  • Contributory infringement is based on the allegation that Amazon Rekognition is a "non-staple article of commerce that has no substantial use" other than in an infringing manner (Compl. ¶23, ¶32).
• Willful Infringement: Plaintiff requests a finding of willful infringement (Compl. p. 10, Prayer for Relief ¶B). The complaint alleges Defendant has "full knowledge" of the patents but does not specify whether this knowledge was obtained pre-suit or post-suit, providing a minimal factual basis for the allegation beyond the notice provided by the lawsuit itself (Compl. ¶22, ¶31).

VII. Analyst’s Conclusion: Key Questions for the Case

• A central issue will be one of technical evidence: Can the Plaintiff, through discovery, obtain and present sufficient evidence to demonstrate that the accused "Amazon Rekognition Face Liveness" service—a third-party, black-box system—in fact performs the specific, multi-step analytical processes required by the asserted claims?
• A key legal question will concern divided infringement: Given that the claimed methods involve actions by a server (Amazon), an end-user device, and software integration by the Defendant, can the Plaintiff establish that the Defendant directs or controls all steps of the method as required to prove direct infringement against it?
• The contributory infringement claim will turn on a critical question of market reality: Is the "Amazon Rekognition Face Liveness" service a staple article of commerce with substantial non-infringing uses, or is it a specialized component whose primary purpose is to perform the patented method?