namisgr

Except that I'm done. I've voiced my opinions to the original question posed, provided rationale with clear support from the relevant fields. I used personal experience with a multiplicity of biological imaging technologies and from having colleagues who use robotics coupled with imaging to perform complex and life saving surgeries, including endovascular and prenatal. I considered the approach for machine learning in this instance as analogous to training new graders, a process that is done all the time at CGC. There comes a point where it's not in my benefit to continue. My opinions from this point forward will have to speak for themselves, as I'm not about to be browbeaten over them, particularly by someone operating outside their sphere of knowledge. I agree to disagree.

This is an exaggeration, as we differ in prior posts in the thread by some margin on the anticipated length of time it may take for the technology to become cost effective. Perhaps it's the result of you mistakenly believing it's going to take technological innovation of NASA proportions to accomplish, when in fact it would be a relatively straightforward undertaking in comparison to the many thousands of examples of machine learning being developed and implemented in the sciences, coupled with robotics and imaging technologies that are highly sophisticated and existing already, requiring only customization for their brand new purpose.

A key word search of PubMed for the term 'machine learning' finds 136,885 peer reviewed publications in professional scientific journals. There are over 8,000 published studies archived for the subfield of 'nervous system' alone. So despite not being easy, it has become routine for those versed and skilled in the requisite programming and computer science aspects. https://pubmed.ncbi.nlm.nih.gov/?term=machine+learning&sort=date&size=20 To boot, the machine learning for grading of comic books is a straightforward application, and made even easier by the codified and systematic process by which numerical grades are derived by professional graders, and the way the codified and systematic processes are used to rapidly train new graders.

Thanks, Mike. This is a promising statement toward making things right with collectors duped by the holder tampering and the rest of us wanting both reliability and accountability from CGC.

Nice catch. The ASM 252 has a color breaking corner crease that is plainly evident at the lower right corner of the front cover, and that would certainly preclude a 9.8 grade so long as it isn't from Shaken Comic Syndrome.

Maybe so. But I still think that the best way to train an algorithm to grade would be to model the training after how new graders are taught to grade.

I just don't think the grader's notes that are recorded and archived are of sufficient detail and reliability for training a machine learning algorithm to grade. Such training would depend on the assessment by each grader of every defect and a structured outline of how they combined for the derivation of their assigned overall grade for each book. I would imagine this is the process by which new graders are trained, not off the often sketchy and incomplete notes routinely recorded, but from one on one discussion of the collection of defects and systematic nature of the grading method between grading trainer and grading trainee.

Same here. When I began collecting Silver Age comics in decent shape a decade before the advent of CGC, I found that VF and VF+ books hit my sweet spot. Really nice looking books for the most part, and much, much better preserved than the majority of Silver Age comics on the market in general, these books were also pretty affordable for a run collector to accumulate. Nowadays, a collector might be better off buying Silver Age VF and VF+ non-keys raw, as the encapsulation process adds significantly to the cost. For bronze age non-keys, I prefer raw books in slightly nicer state of preservation, with VF/NM and NM- commonly being a good trade-off between appearance and cost. But as one goes down the line to books of lower and lower grades for either era, the cost hit from slabbing becomes more and more acute. Of course for those able to afford higher grades than 8.5, their budgets and their collecting esthetic probably determine their 'ideal' grade.

You can train an amateur in comic book grading to become a professional comic book grader at CGC. The training is iterative, and almost certainly takes months rather than years Whoever thinks you cannot train a machine learning algorithm to become a comic book grader with accuracy comparable to many of the professional graders themselves is either unaware of what machine learning algorithms have been accomplishing over the past decade, unknowing about how iterative training refines and improves the accuracy of the algorithms over time, or both. Start with a grading algorithm based on the systematic methodology CGC graders use to derive final structural grades. Have it grade a learning set of books, for purposes of discussion let's say a hundred that span the full range and severity of defect types, and all of which have also been graded and thorough grader's notes recorded by five CGC graders each working independently. Have the finalizer assign a final grade to each book. The algorithm will then incorporate information culled from the ranges, disparities, and detailed grader's notes to derive a modified protocol that better fits the human-derived grades for the learning set. Then the entire process will be repeated. After each repeat, the algorithm can be expected to come closer and closer to the human-derived grades, and especially the final grade. Finally, the algorithm and the team of professional graders will evaluate a test set of a hundred different books, so that its accuracy in a real world-relevant scenario may be assessed, and compared with each of the graders who were not the finalizer. The image capturing component of automated grading of this type is technology that is not 'Artificial Intelligence', and is pre-existing. This general approach has already been used to devise algorithms that can translate ensembles of nerve cell impulses from the region of the neocortex that directs leg movements in paralyzed persons with damaged spinal cords into electrical outputs that stimulate leg movement-controlling nerve cells in the spinal cord below the site of injury, allowing a person previously in a wheelchair and paralyzed for over a decade to use their thoughts and a brain-spinal cord interface machine with stimulating electrodes and an algorithm that decodes the nerve cell activity to allow them to initiate and engage in walking. Given the systematic nature of the professional comic book grading process, and the relative simplicity of the machine learning required, it would be much, much easier to devise an accurate grading algorithm than it would to restore limb movements, or vision, or hearing, or speech in people who've lost these functions to injury or disease, all of which have been accomplished over the past five years. Making such an automated grading system cost effective is a separate matter, and may certainly take awhile. But I am of the belief that the technological aspects could be readily achieved.

Some of the perp's information will have to be shared for the searching to continue for slabs with bogus grades and for consumers to be able to evaluate the slabs in their collections and be made aware of suspect slabbed books going forward. Names/aliases and city locations used by the scammer to sell books at various venues must be spread clear and wide, at a minimum. And lists of certification numbers from all submissions to CGC by the perp must be compiled and released.

I see a lot of potential 9.8s in that picture.

I agree that a fully satisfactory response from CGC would be to offer compensation for the lost value of books that, upon re-grading, are found to be of considerably less value owing to a change in either the numerical grade, the book specifics, or both. But it, too, starts with CGC agreeing to take back any comic submitted by the perpetrator and regrading it for nothing. As to why anyone would resubmit a book that might be found out as being tainted, it would be for its sale value. Just as Jason Ewert submissions instantly cratered in marketplace value, so too will any slab with a certification number tied to the fraudster.

One thing the company plainly should do, which they did with all Jason Ewert submissions, is accept back any and every slabbed book submitted by the perpetrator, and re-grade them free of charge. I can't see any measure less than that being satisfactory.

And indeed, one or more people at CGC may very well be working on the matter as we converse. But I would be surprised if a thorough response involving many people is forthcoming before the new year, as the issue is too important for the involvement of just a couple of higher ups.

Xylene is used routinely in histology, the microscopic preparation and examination of cells and tissues. I had 35 years of inhaling the stuff , but in very, very small quantity since work with it was always performed and only allowed under a protective fume hood. It's indeed a near-universal dehydrating agent and clears/dissolves plastics. It's not what I would call 'sweet' smelling either. It's unpleasant. And besides the acute problems that exposure to it causes that you listed, it can also cause chronic problems, including damage to the liver, kidneys, and reproductive organs. Nasty stuff, and long classified by OSHA as a biohazard. Very familiar base structure to any chemist, and usually referred to as xylenes, since the stuff is commonly a mix of three different isomers:

While I'm far from an apologist for the company, I think CGC is on holiday and so understand why there wouldn't be a formal response. I'd expect one after the first of the new year.

This was another instance when it was all about the front cover.

Sounds like a patient friendly advance. But for comic books, very little three dimensional reconstruction is even necessary. The pages themselves wouldn't require it, just a head on image and a couple taken at slightly non-perpendicular angles, which would pick up any bends, tears, pieces out, dents, soiling, rust, siamese pages, and the like with ease, while also generating a page quality score for each page that could be compiled into a final grade. The imaging of each exterior and interior cover page would require several angled views in addition to the perpendicular one, and possibly to a stack of multiple perpendicular views acquired at different z-planes, to detect surface impressions, dents, waves, and other imperfections of paper flatness. It's technology reminiscent of what's been available in microscopy and medical imaging for decades now for deriving three dimensional reconstructions from series of what are essentially two dimensional images. The photographic and image reconstruction methods are all automated and very fast.

Page turning device that's just a wee bit simpler than the work NASA or an endovascular or prenatal surgeon does: And, to head off any suggestion at the pass, the idea of posting this video is not to suggest that this device could be used as part of an automated comic book grading system. Instead, the idea is to illustrate how little sophistication and cost would be involved in developing a system sufficiently gentle, reliable, programmable, and cheap to turn pages and partner with an imaging and book positioning system to readily capture all the images required for an algorithm developed by machine learning from the output of professional CGC graders to devise an automated grading system.

It's listed on the Exchange, in which case CLink would not have the book in hand and the ability to assess it until the book sells, when the seller must send it to the Link. I wonder if they've unknowingly auctioned any of the fraudulent books?

I said nothing of the kind, and I don't agree with the statement. Like I wrote already, there have been over the past several years repeated demonstrations applying brain-computer interfaces and machine learning to derive algorithms and perform functions previously lost in humans through nervous system injury or disease. People paralyzed have walked under their own power, "spoken" language via computer at a near normal rate, and achieved restored auditory and visual functions, all of which they were unable to do for up to a couple of decades prior. To develop a computer-based method for grading comic books with the accuracy of professional human graders would be, comparatively, far easier to do than the achievements already made with defective nervous systems. After all, there already exists a systematic framework graders use on every book that takes into account the number, severity, nature, and placement of wear and defects that reduce numerical grade, that was implemented to maintain consistency over time and among different graders. Also like I wrote already, the robotics technology for turning comic book pages and capturing their images, including the outside and inside covers at multiple angles, is pre-existing and I see no reason why it would be prohibitively expensive. Finally, as I wrote already, the capacity to achieve economic equivalence with the current all human grading approach is, in my opinion, not as far away as you seem to believe, and might be cost effective in as little as a decade, and I wouldn't rule out even shorter. The emergent computer and robotics technologies to perform these tasks could work around the clock and on weekends, and there could be multiple setups operating at once. The upfront financial burden of implementation resolves over time into a far, far smaller set of maintenance expenses.

Found it:

Gilbert Shelton!

Seeing the differences in Forum reactions to and discussions about the two films is an observation, and not an implication. Having tired of being browbeaten, ciao.