PREVIOUS ADVICE 
In your column, you frequently mention the importance of making contact with program officers (POs). But what if the program officer assigned to your NIH grant application does not respond to phone calls or e-mails? I was in this situation recently when I got my critiques back and tried to call and e-mail the PO that was assigned to me for advice. I could never get hold of him. I always just got through to his voice mail, he never returned my phone calls, and he never responded to my e-mails, except for one time when he said he would call me the following day. He didn't. I realize that these people are busy, travel a lot, and probably have a hundred people like me trying to get their advice. However, knowing that doesn't do me any good.
This puts me in a difficult position: Do I keep pestering the PO and his staff day after day, until I finally make contact, with the risk that he'll get angry and thus not be very helpful? Or do I just give up? Thanks for any advice you can provide.
You're right: I often tell people to call their program officer. Why do I say that? Mainly because every time I talk to one of the higher-ups at NIH, that's exactly the advice they tell me to pass along. And it makes good sense: No one knows an NIH program better than the program officer in charge.
So why wouldn't your program officer return your calls and e-mails? I posed that question to two of my favorite knowledgeable sources, and I got two quite different answers. The first knowledge source (a long-time PO who I will hereafter refer to as KS1, for "knowledgeable source #1") assumed that you must have called at a bad time, within a few weeks of a submission deadline. At those times, said KS1, POs can be very, very busy.
My 2nd knowledgeable source (KS2), a current NIH insider, pointed out that there are other reasons why that PO isn't returning your calls. "There is a lot of variability," he suggested. "Some POs are very accessible; others like to hide behind their Web site. I do not believe it is a function of institutional culture, rather of the individual." Some folks just aren't as good at getting back to you as other folks. Fair enough.
Furthermore, there may be other reasons--other than pending deadlines--why a program officer might be busy, postulated KS2. "It is certainly true that everyone is under a lot of pressure these days," he said. "Between outsourcing, centralization of support staff functions, and near-paralysis of hiring, everyone ends up doing what used to be someone else's job in addition to his own. Also, with tight paylines [i.e., more rejections] more investigators need help." More rejections means more questions, and more conversations with scientists who are angry or distraught. Your tax dollars hard at work.
But never mind why: What should you do about it? What should you do if the PO doesn't call you back? That's a tough one. One thing you definitely should not do is risk alienating your PO; that relationship is too important. A situation like this must be handled with great delicacy. Here is the approach I would take, drawing on the advice of KS1 and KS2:
Try again, this time sending an e-mail asking if it would be possible to set up a meeting to discuss your proposal. Be formal and pointedly respectful but not ingratiating. The goal is to establish the foundation for a good, mutually respectful working relationship for the long haul. Allow several days for a reply.
If after several days you haven't heard anything, call. If the PO answers, respectfully request a meeting at some later, mutually convenient date and time. If an assistant answers, ask if it would be possible to set up a meeting with the PO. If the assistant says no, politely request that your phone call be returned at the PO's soonest convenience.
If, after a couple more days, you've still heard nothing, locate another PO who works in a related field. "If you're applying for a fellowship," suggests KS1, "find a PO in the same area who is in charge of R01s." Request a telephone meeting. During your conversation--don't do this via e-mail--you might inquire after the health of your PO, noting that you've found him or her difficult to get in touch with lately. Ask your questions. The reply, very likely, will be, "you really should talk to Dr. X." Thank this person very much.
That ought to do it. Do this, and I'll almost guarantee a call back. Let me know how it turns out.
Thank you for your service to the community. I find your advice helpful! It seems that most of the questions/answers are targeted toward biological and basic science areas, so I decide to ask you my specific question. I am starting as a junior faculty this fall in UIUC. My area of research is computer vision (primary) and human vision. Computer vision crosses the fields of engineering, information science, and machine learning. I know that in general the traditional grants in the field are CISE (NSF) and DARPA. Could you please help me to list out a more comprehensive list of grant sources?
Thank you very much!
Times are tough for computer scientists, and getting tougher. Ever since the dot-com boom went bust, private-company support for computer science research has dried up, leaving NSF to support the great majority of non-defense CS research. The Department of Defense is a major player, as you have noted, but they have started to channel more of their resources into private sector contractors and to put more security restrictions in place that are not friendly to foreign nationals. I'm not sure how long DARPA will remain a useful source of funding for academic computer scientists.
The key to finding funding in a climate like this is to do some thinking about the potential applications of your research. While it's usually a mistake to allow funding opportunities guide your research, thinking hard and creatively about the overlap between your interests and the missions of the various funding agencies can be a useful exercise. Given your interest in both computer vision and human vision, that thinking ought to lead you to the National Institutes of Health, and--in particular--the National Eye Institute.
For a physical scientist, NIH can be daunting. Compared to NSF, the system is large and complex. That is why we wrote Getting an NIH R01 Grant . It's the fastest way to get up to speed on the complexities of NIH.
NSF's CISE program remains an excellent choice. With a little imagination, your research might be made to fit the mission  of the Air Force Office of Scientific Research. Similarly, you might manage to get the folks at the Army Research Office interested in research like yours; here's a link  to their broad agency announcement. Finally, it is possible that your research could find a home within one of NASA's instrumentation programs.