iNPUT-ACE video evidence software and training

The People v. Hung Tran

Around 2:00 am on August 27, 2016, the Gaslamp Quarter in San Diego was crowded with people enjoying the nightlife. Two groups inside the Gaslamp Café had been quarreling and were asked to leave the restaurant. One group consisted of Hung Tran and several of his friends and relatives. The members of the group on the other side of the conflict were unidentified. There was no known gang affiliation for anyone involved in the argument.

Once outside, the yelling continued and eventually escalated into physical altercations.

At the nearby New Yorker restaurant, two good Samaritans, M.C. and his friend, had just finished eating some pizza and observed the confrontation. They stepped in to help break up the fighting, but the members of Hung Tran’s group instead turned on M.C. and his friend. M.C. was body-slammed to the ground and landed on his head. Several members in Tran’s group then continued to punch and kick M.C. on the ground. The attack left M.C. a quadriplegic.

Hung Tran was convicted by a jury of “assault by means likely to produce great bodily injury (count 1) and mayhem (count 2)”, and he was sentenced to four years in prison. (See endnote)

Tran appealed on several grounds. He contended that the videos produced by the prosecution’s forensic video expert were “doctored” and that the court admitted them into evidence erroneously. He also contended that there was not sufficient evidence to support his conviction, that the court erroneously admitted lay opinion testimony from a witness, and that his trial counsel was prejudicially ineffective. After review, the Court of Appeal determined that none of Tran’s claims had merit and it affirmed the conviction.

The prosecution’s video expert was Grant Fredericks, who recently delivered an excellent training session during the iNPUT-ACE sponsored Online Video Evidence Symposium in June. Fredericks is a certified Forensic Video Analyst with over 30 years of experience conducting video-centric investigations and analysis. He has testified hundreds of times in criminal and civil trials (equally as a prosecution and defense expert) and has successfully testified during several Daubert and Kelly-Frye motions. In the case of The People v. Tran, Fredericks testified as a forensic video expert for the prosecution. In the appeal, Tran’s lawyers challenged Fredericks’s evidence, arguing that Fredericks “became the [fact finder] who decided and shaped the facts of the case instead of the jury.” Tran contended that Fredericks “improperly enhanced” the videos (referencing height/width ratio adjustments, synchronization, blur correction, and the addition of color-coded arrows), and that these enhancements “resulted in Fredericks invading the province of the jury by making key factual determinations.” Tran further argued that Fredericks’ enhancement techniques should be evaluated under the Kelly-Frye test.

The court summarily found that Tran’s challenges to the video lacked substance. However, because the admissibility of this kind of evidence was not addressed in any published California case, the court opted to specifically elaborate on the issue.

In order to validate the Kelly-Frye test, the appellate court reviewed Mr. Fredericks’ methodology. Fredericks explained that he first interrogated each file thoroughly and analyzed compression levels, frame rates, and pixel matrixes. His primary tool was iNPUT-ACE, which the court noted was software that “allows investigators to process, enhance, and analyze videos, including movement management.” Fredericks testified that the procedures he employed are “currently [used by] several hundred certified analysts and technicians, all of whom have been trained in this process.”

Fredericks then described how he accurately synchronized each of the videos in order to create chronological sequences to view multiple videos on a single screen. Once synchronized, he was able to track individuals from one video to the next and apply color-coded arrows. The synchronization was critical to making sense of the commotion, constant movement, and different camera perspectives. This side-by-side synchronization allowed Fredericks to continue to track individuals even through camera angles where the image may have been otherwise too blurry or insufficient for identification on its own. The trial court agreed that, without Fredericks’ work, tracking the subjects would have been impossible.

[Author’s Note] The value of synchronizing the video can be a critical factor in tracking the movement of targeted individuals and in helping the trier of fact to comprehend the timeline. Often, simply playing a series of videos in a linear or chronological order can cause more confusion than clarity. What Fredericks did was synchronize multiple videos onto a single screen to help the trier of fact understand how each person moved and their relationship to one another from camera to camera. To add to the complexity, integrating multiple videos from different sources and tagging each individual was a key element to helping the jury understand who to follow with their eyes and when.

During our Training Symposium, Certified Forensic Video Technician Brandon Wahl engaged in a detailed discussion on how to address multiple sources of surveillance video. One of the most common challenges he explored is that the date/time stamps from DVRs are often incorrect. This means that investigators must find the timing offsets in order to track a suspect’s activity from one system to the next. He explained the best practices for determining those timing offsets and how to perform date/time synchronizations within iNPUT-ACE. Once all systems have been synchronized, investigators can snap to any point in time and see all of the activity from all camera angles to be able to track activity more easily.

With regard to reproducibility, Fredericks explained how he carefully documents all of his work so that any similarly trained peer could recreate each step, review, and test. He specifically noted that there is a saved record of all markers and of every portion of his work product for peer review. As we discussed in the previous blog from this series, Kelly-Frye, Daubert, Mohan, and Why You Need to Understand These Cases, it is critical for an expert’s work to be reproducible. This is why iNPUT-ACE provides the ability to save Projects and Workflows so that you can reload all of your work and produce a report containing all the actions that you applied to any of your videos.

[Author’s Note] As a former law enforcement agent and retired Police Chief, I feel this next section is relevant for all practitioners. For up-and-coming experts, it is a critical element. For experienced practitioners, it is a great refresher.

In a California Evidence Code section 402 hearing, Fredericks defined the limits of his area of expertise and made it clear that he would not “opine in various areas that might infringe on the province of the jury.” He explained this means that, while he would describe the X/Y coordinates specific to an action, he would not testify as to whether or not the video showed contact between people. “I’m not a use-of-force expert. I am not going to tell the jury, ‘here, you should see this person kicking this person and this person punching that person.’ That’s not what I am going to do.” While M.C.’s injuries were at issue in this case, Fredericks remained firm that he “wouldn’t be offering any opinion, statement, comment about the nature of [M.C.’s] fall and what injuries were caused” or “who caused the motion to the ground.”

Fredericks went on to explain the necessity of his expertise. The average person is often unable to comprehend when different videos capture the same people or activity but from a different angle. For example, in crossed-axis tracking, a person may appear to move from left to right in one camera view, but a camera on the opposite side of the street will show that same person as moving from right to left. A layperson may also be ill-equipped to comprehend the significance of blurring (which can represent fast motion) and the effects of infrared on colors (which can cause a person to appear to be wearing different clothing in videos of the same event). Fredericks countered allegations that he “altered” the videos by explaining that, while he did add color-coded arrows, “the meaning of all the videos are maintained” and “the actual images are completely untouched.”

Following Fredericks’ pre-trial hearing testimony, the trial court concluded that Fredericks’ testimony was “quite necessary” and “related to a subject that is sufficiently beyond common experience.” The court noted that the videos were “busy, full of commotion, hard to follow, active, and even with the benefit of the colored indicators, . . . are very difficult to track.” Fredericks was allowed to testify at trial and the videos with color-coded arrows were admitted into evidence.

The trial court ruled that there was no need to apply to the Kelly-Frye test since none of the methods employed by Fredericks were novel or new. The process of synchronization in particular was described as using “no special method” and instead “rests on common sense.” Fredericks was able to explain his steps so effectively during cross-examination that the court was confident that “the jurors are going to very easily not be blinded by any of the science. There really isn’t science to this.”

I will close with an opinion that we hope creates a positive discussion amongst industry experts. Many experts have a desire to call everything scientific, but that can lead to unnecessary challenges at trial. A process such as putting an arrow over someone’s head and tracking them isn’t a hard “science”, but as the trial court noted, it was very difficult for the trier of fact to follow the movements of the individuals even with the arrows and “impossible” without Fredericks’ expert assistance.

Fredericks’ expertise was needed for the jury to properly comprehend what was depicted in the videos. That’s what a forensic video analyst does: accurately interpret video evidence and make it easier for the trier of fact to understand. In this matter, the analyst’s work didn’t require the application of a hard “science” in order for the evidence to be admissible.


The full text of the appellate court’s opinion can be found here:

More From Our Series – FVA Challenges

iNPUT-ACE video evidence software and training

Kelly-Frye, Daubert, Mohan, and Why You Need to Understand These Cases

In our line of work, it is common to hear comments regarding Kelly-Frye, Daubert and Mohan challenges. At iNPUT-ACE, our users frequently ask us what these references mean in regard to legal challenges of video evidence…