FaceTec Inc v. Jumio Corp

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: FaceTec, Inc. (Delaware)
  • Defendant: Jumio Corporation (Delaware)
  • Plaintiff's Counsel: One LLP
• Case Identification: 5:24-cv-03623, N.D. Cal., 06/14/2024
• Venue Allegations: Venue is alleged to be proper based on Defendant Jumio Corporation having a regular and established place of business in the Northern District of California, specifically in Sunnyvale.
• Core Dispute: Plaintiff alleges that Defendant's biometric security software infringes four U.S. patents related to using perspective distortion from 2D images to verify the liveness and three-dimensionality of a person.
• Technical Context: The technology at issue involves software-based methods for biometric liveness detection, designed to prevent spoofing in online identity verification without requiring specialized 3D cameras or hardware.
• Key Procedural History: The complaint alleges a complex history between the parties, stating they had a prior contractual relationship that Defendant terminated before contracting with a competitor, iProov Ltd. Plaintiff previously sued iProov in the District of Nevada in late 2021 for infringement of two of the same patents asserted here. Plaintiff also alleges it sent a cease-and-desist letter to Defendant in August 2023, putting Defendant on notice of the alleged infringement prior to this lawsuit.

Case Timeline

Date	Event
2014-08-28	Earliest Priority Date for '471, '606, '938, and '910 Patents
2020-09-15	U.S. Patent No. 10,776,471 Issued
2021-04-07	Jumio announces partnership with iProov Ltd.
2021-10-26	U.S. Patent No. 11,157,606 Issued
Late 2021	FaceTec files suit against iProov Ltd. in D. Nev.
2023-07-04	U.S. Patent No. 11,693,938 Issued
2023-08-23	FaceTec sends cease and desist letter to Jumio
2024-01-16	U.S. Patent No. 11,874,910 Issued
2024-06-14	Complaint Filed

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

U.S. Patent No. 10,776,471 (the '471 Patent) - "Facial Recognition Authentication System Including Path Parameters"

• Patent Identification: U.S. Patent No. 10,776,471, "Facial Recognition Authentication System Including Path Parameters," issued September 15, 2020.

The Invention Explained

• Problem Addressed: The patent addresses the insecurity of conventional two-dimensional (2D) facial recognition systems, which can be "tricked" or spoofed using photographs or video displays of a user's face Compl. ¶8 '471 Patent, col. 1:61-67
• The Patented Solution: The invention proposes a method to verify a user is a live, three-dimensional (3D) person by using a standard 2D camera to capture images at two or more different distances. The system then analyzes these images for "perspective distortion"-natural changes in the relative proportions of facial features (e.g., the nose appearing larger relative to the rest of the face at close range) that occur when a 3D object is viewed from different distances Compl. ¶¶12-13 '471 Patent, abstract '471 Patent, col. 28:45-67 The presence of this expected distortion is used to authenticate the user as a live human.
• Technical Importance: This approach provides a method for robust 3D liveness detection that works on standard computing devices (smartphones, laptops) without requiring specialized hardware like infrared or 3D cameras Compl. ¶17

Key Claims at a Glance

• The complaint asserts at least independent claim 10 Compl. ¶35
• The essential elements of Claim 10 include:
  • Capturing a first image of the user at a first distance.
  • Processing the first image to create first data.
  • Moving the camera or user to a second, different distance.
  • Capturing a second image at the second distance.
  • Processing the second image to create second data.
  • Comparing the first and second data to determine if "expected distortion" exists.
  • Authenticating the user when the expected distortion is present.
• The complaint does not explicitly reserve the right to assert dependent claims for this patent, but refers to "Claim 10 and other claims" Compl. ¶35

U.S. Patent No. 11,157,606 (the '606 Patent) - "Facial Recognition Authentication System Including Path Parameters"

• Patent Identification: U.S. Patent No. 11,157,606, "Facial Recognition Authentication System Including Path Parameters," issued October 26, 2021.

The Invention Explained

• Problem Addressed: The patent targets the same problem as the '471 Patent: the vulnerability of 2D facial recognition systems to spoofing attacks from non-live artifacts like photos Compl. ¶8 '606 Patent, col. 1:62-67
• The Patented Solution: The patented method verifies a user's three-dimensionality by capturing images at two different distances, processing them into first and second biometric data, and then performing two comparisons. First, it determines the biometric data sets do not match. Second, it determines that the differences between the data sets do match the expected differences caused by the change in distance (i.e., perspective distortion). A positive 3D verification occurs when both conditions are met '606 Patent, abstract '606 Patent, claim 1
• Technical Importance: The invention provides a hardware-independent software solution for liveness detection, enabling its use on a wide array of existing consumer devices Compl. ¶17

Key Claims at a Glance

• The complaint asserts at least independent claim 1 Compl. ¶48
• The essential elements of Claim 1 include:
  • Capturing a first image at a first distance.
  • Processing the first image to obtain first biometric data.
  • Capturing a second image at a second, different distance.
  • Processing the second image to obtain second biometric data.
  • Comparing the first and second biometric data to determine if they match.
  • Comparing the first and second biometric data to determine if differences match expected differences from the change in distance.
  • Determining the user's face is 3D when the data does not match, but the differences are expected.
• The complaint refers to "Claim 1 and other claims" of the patent Compl. ¶48

U.S. Patent No. 11,693,938 (the '938 Patent) - "Facial Recognition Authentication System Including Path Parameters"

• Patent Identification: U.S. Patent No. 11,693,938, "Facial Recognition Authentication System Including Path Parameters," issued July 4, 2023.
• Technology Synopsis: This patent describes a method for detecting spoofing attempts by determining when a user's face does not exhibit three-dimensionality. It involves capturing images at different distances and, if the expected distorting changes are absent, concluding that the subject is not a 3D person '938 Patent, abstract '938 Patent, claim 8
• Asserted Claims: At least independent claim 8 Compl. ¶61
• Accused Features: The complaint alleges that Jumio's liveness detection technology uses this method to identify and reject 2D spoof artifacts Compl. ¶61 Exh. G, p. 25

U.S. Patent No. 11,874,910 (the '910 Patent) - "Facial Recognition Authentication System Including Path Parameters"

• Patent Identification: U.S. Patent No. 11,874,910, "Facial Recognition Authentication System Including Path Parameters," issued January 16, 2024.
• Technology Synopsis: This patent describes a method for evaluating the three-dimensionality of a user by capturing images at different distances. It requires processing the images into data and comparing the data sets to determine if expected differences exist, which would indicate three-dimensionality '910 Patent, abstract '910 Patent, claim 10
• Asserted Claims: At least independent claim 10 Compl. ¶74
• Accused Features: The complaint alleges that Jumio's liveness detection technology uses this method to confirm that a real user is three-dimensional Compl. ¶74 Exh. H, p. 20

III. The Accused Instrumentality

Product Identification

The complaint names the "Jumio Accused Instrumentalities," which encompass a range of products and services for biometric security Compl. ¶28 These include technologies marketed as "Liveness Assurance," "Basic Face Verifier," "Express Liveness," and "Liveness Checks," some of which are noted to be sourced from vendor iProov Ltd. Compl. ¶28.a Compl. ¶28.b

Functionality and Market Context

The complaint alleges the accused products function by prompting a user to position their face at multiple distances from a device's camera, capturing images at those distances, and then comparing the image data to evaluate the user's 3-dimensionality Compl. ¶28.d The complaint includes a screenshot from Jumio's iOS application that prompts the user to "Move closer," illustrating the functionality of capturing images at different distances Exh. F, p. 12 Jumio is identified as a direct competitor to FaceTec in the biometric liveness detection software market Compl. ¶30 The complaint explicitly carves out and does not accuse products that "solely" use "Genuine Presence Assurance" or "Dynamic Liveness," which rely on controlled illumination (flashing colors) Compl. ¶28, fn. 2

IV. Analysis of Infringement Allegations

'471 Patent Infringement Allegations

Claim Element (from Independent Claim 10)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
capturing at least one first image of the user taken with the camera of the computing device at a first location which is a first distance from the user;	The Accused Instrumentalities prompt the user to place their face in an on-screen oval and capture at least one image at an initial distance.	¶28.d; Exh. E, p. 2	col. 27:4-7
processing the at least one first image or a portion to create first data;	The captured image is processed and sent from the user's device to a server, creating or comprising "first data."	Exh. E, p. 4; Exh. E, p. 16	col. 27:8-10
moving the camera...or the user moving...to change the distance...to a second distance;	The system prompts the user to "move closer" or otherwise changes the on-screen oval, inducing the user to change the distance between their face and the camera.	Exh. E, p. 10	col. 27:11-17
capturing at least one second image of the user taken with the camera...at the second distance...being different than the first distance;	The camera captures at least one second image of the user after the user has moved to the second, different distance.	Exh. E, p. 10	col. 27:18-22
processing the at least one second image or a portion thereof to create second data;	The second captured image is processed to create "second data" on a remote server.	Exh. E, p. 16	col. 27:23-25
comparing the first data to the second data to determine whether expected distortion exist...which indicated three-dimensionality of the user;	A "Jumio Neural Network" allegedly analyzes and compares the first and second data to identify expected distortion that indicates the user is 3D.	Exh. E, p. 18	col. 27:26-30
authenticating the user when the differences between the first data and the second data have expected distortion...	The system authenticates the user upon confirming the presence of expected distortion between the data sets.	Exh. E, p. 23	col. 27:31-38

'606 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
capturing at least one first image...at a first distance from the user;	The Accused Instrumentalities capture a first facial image at an initial distance, prompted by an on-screen guide.	Exh. F, p. 2	col. 45:18-22
processing the at least one first image to obtain first biometric data...;	The captured image is processed, and the resulting biometric data is sent to a server for analysis. A diagram in the complaint shows this data flow (Exh. F, p. 5).	Exh. F, p. 4	col. 45:23-25
capturing at least one second image...at a second distance...being different than the first distance;	After prompting the user to "move closer," the system captures a second image at a new, different distance.	Exh. F, p. 10	col. 45:26-31
processing the at least one second image to obtain second biometric data...;	The second image is processed to obtain second biometric data, which is also sent to the server.	Exh. F, p. 16	col. 45:32-34
comparing the first biometric data with the second biometric data to determine whether the first biometric data matches the second biometric data;	The system's "Jumio Neural Network" is alleged to internally compare the data sets to identify differences.	Exh. F, p. 19	col. 45:35-38
comparing the first biometric data to second biometric data to determine whether differences...match expected differences resulting from movement...;	The neural network allegedly analyzes whether the identified differences correspond to the expected distortion from a 3D face imaged at two distances.	Exh. F, p. 19	col. 45:39-47
determining that the user's face is three-dimensional when: the first biometric data does not match the second biometric data; and the second biometric data has the expected differences...	The system determines the user is 3D when the data sets are different but the differences are the expected ones, confirming liveness and rejecting 2D spoofs.	Exh. F, p. 24	col. 45:48-59

• Identified Points of Contention:
  • Technical Questions: A central evidentiary question will be what the accused "Jumio Neural Network" actually analyzes. Does the complaint provide sufficient evidence that the system specifically detects "perspective distortion" as taught by the patents, or does it rely on a different liveness detection technique that may not fall within the claims' scope? The complaint relies on observations of the user interface and public marketing materials, which may not definitively prove the underlying technical mechanism.
  • Scope Questions: Claim 1 of the '606 Patent includes the step "determining that the user's face is three-dimensional when: the first biometric data does not match the second biometric data." The interpretation of "does not match" will be critical. It could be read to mean the raw data is simply different (as would be expected with any change in perspective), or it could be argued to imply a specific type of mismatch test. The precise function of this limitation will raise questions for claim construction and infringement analysis.

V. Key Claim Terms for Construction

For the '471 and '606 Patents

• The Term: "expected distortion" ('471 Patent) / "expected differences" ('606 Patent)

• Context and Importance: This concept is the technological core of the asserted patents. The infringement case hinges on whether the accused products detect the specific type of "expected" visual changes that the patents describe as indicative of a live, 3D person. Practitioners may focus on this term to dispute whether the accused system's analysis aligns with the patent's teachings.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The specification describes the phenomenon in general terms, such as the "fish-eye effect" becoming "more pronounced" at close range, causing "relative dimensions of the person's face appearing different" ('471 Patent, col. 28:45-52). This language may support a construction covering any algorithm that detects general changes in facial proportions due to distance.
  • Evidence for a Narrower Interpretation: The specification provides a highly specific example: "a person's nose may appear as much as 30% wider and 15% taller relative to a person's face when the image is taken at a close proximity as compared to when the image is taken at a distance" '471 Patent, col. 28:52-57 A defendant may argue this language limits the claims to methods that detect this specific type and magnitude of nasal distortion, rather than any general change.

• The Term: "biometric data" ('606 Patent)

• Context and Importance: The claims require processing images to "obtain biometric data" and then comparing that data. The definition of this term will determine what form the data must take and what kind of comparison is performed, which is central to the infringement analysis.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The specification provides a broad definition: "The term biometric data may include among other information biometric information concerning facial features (distorted or undistorted) and path parameters" '606 Patent, col. 6:40-43 This could be argued to encompass a wide range of data types, from raw pixel information to feature vectors.
  • Evidence for a Narrower Interpretation: The claims describe a multi-step process of "processing" an "image" to "obtain" "data," which is then "compared." A defendant may argue this requires a distinct transformation from image to a specific structured data format (e.g., a numerical map of facial features) before comparison, rather than a direct comparison of image files or raw pixels.

VI. Other Allegations

• Indirect Infringement: The complaint alleges both induced and contributory infringement. Inducement is based on allegations that Jumio provides on-screen prompts (e.g., "Move closer"), user manuals, and other instructional materials that actively guide end-users to perform the patented methods Compl. ¶37 Compl. ¶39 Contributory infringement is alleged on the basis that Jumio's software constitutes a material part of the invention, is not a staple article of commerce, and is especially designed for an infringing use Compl. ¶36
• Willful Infringement: The complaint alleges willfulness based on both pre-suit and post-suit knowledge. It claims Jumio had pre-suit knowledge due to its status as a direct competitor, a prior contractual relationship with FaceTec, and a cease-and-desist letter sent in August 2023 that specifically identified the asserted patents and accused products (Compl. ¶30; Compl. ¶31; Compl. ¶32). Continued alleged infringement after the filing of the complaint forms the basis for post-suit willfulness.

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

• A core issue will be one of technical operation: what is the specific mechanism by which Jumio's "Liveness Detection" software functions? The case will require evidence beyond user-facing prompts to determine if the system's "neural network" in fact performs the analysis of "perspective distortion" as required by the patent claims, or if it relies on an alternative, non-infringing liveness detection method.
• A second key question will be one of definitional scope: how will the court construe the term "expected distortion/differences"? Whether this term is interpreted broadly to cover any algorithm that detects changes in facial proportions due to distance, or narrowly to require detection of the specific types and magnitudes of distortion described in the patent's examples, will be critical to the infringement outcome.
• A final central question will relate to intent and damages: given the detailed history of prior business dealings, a lawsuit against Jumio's technology partner (iProov), and a pre-suit notice letter, a significant focus of the case will be whether Jumio's alleged infringement, if found, was willful, which could expose the company to enhanced damages.