NIELSEN COMPANY (US), LLC v. TVISION INSIGHTS, INC.
Authored by: Jeremy J. Gustrowsky
The Federal Circuit recently affirmed a Patent Trial and Appeal Board decision finding claims of Nielsen’s U.S. Patent No. 11,470,243, directed to audience measurement using cameras, unpatentable as obvious. The case centered on whether a scientific publication on facial expression analysis qualified as analogous prior art to a patent aimed at counting and identifying television viewers.
The ‘243 patent describes an audience measurement device that uses a camera to capture images of viewers. Low-resolution images are used to count people and detect head orientation, while higher-resolution images are used for facial recognition against a database of facial signatures. TVision Insights petitioned for inter partes review, relying on a publication by Ying-li Tian titled “Evaluation of Face Resolution for Expression Analysis,” combined with other references including U.S. Patent Application Publication Nos. 2002/0059577 (Lu), 2010/0274372 (Nielsen-372), and 2010/0066822 (Steinberg). The Board found the challenged claims unpatentable, and Nielsen appealed.
Nielsen’s first argument was procedural, contending that the Board violated the Administrative Procedure Act by finding Tian to be “reasonably pertinent” analogous art when TVision’s petition had argued only that Tian was in the “same field of endeavor.” The Federal Circuit rejected this on three independent grounds. The two analogous art prongs often overlap, and the reasonably pertinent theory was implicit in the petition. Nielsen actually addressed both prongs in its Patent Owner Response, and Nielsen could not identify any prejudice from the alleged lack of notice, making any error harmless.
On the merits of the analogous art question, Nielsen argued the Board should have limited its analysis to the specific problems expressly identified in the specification, namely drawbacks associated with frequent activation of illumination sources like power drain, heat, and audience annoyance. The Federal Circuit disagreed, citing KSR for the proposition that a patent may address multiple problems and that “the objective reach of the claim” governs, not the patentee’s stated purpose. Since the challenged claims did not even recite a light source, limiting the relevant problem to illumination concerns would be inappropriate. The Board properly identified the broader problem as image processing and facial detection, supported by the patent’s title, background, and claim language.
Because Tian explored facial detection at reduced resolutions, sharing a purpose with the ‘243 patent, it “logically would have commended itself” to an inventor considering the same problems. On the obviousness merits, the court rejected Nielsen’s attempt to read unstated limitations into the claims. Claim 4 requires “reduc[ing] a resolution” but is not limited to the specification’s pixel “binning” technique, so Tian’s down-sampling suffices. Claims 5 and 6 do not require that only the first image be reduced in resolution, nor that the second image be full-resolution. The Board properly credited expert testimony that reducing processing time is a well-known motivation in face detection, supporting a motivation to combine the references.
The decision reinforces two practical points for patent practitioners. First, arguments about analogous art should not be tethered too tightly to problems expressly recited in the specification, since the claims themselves and general problems addressed by the invention can define the relevant inquiry. Second, when drafting claims, be aware that broad claim language can invite a broad framing of the “problem” for analogous art purposes, potentially bringing in prior art from adjacent technical fields that the applicant may not have anticipated.