I have to do my defense seminar in couple days. I personally hate to kick off the talk with "hello and thank you for attending my talk". I am wondering if you know better ways/phrases to start a talk. You attended many conferences so I bet you saw some good starts.
Start it with a compelling question that add some context to your talk. A recent talk at a conference I attended started like this:
What is the largest sumfree subset of the numbers 1 to n?
This is a well-known problem first posed by Erdos, and if someone can say that their research leads to a nice way to solve it, a lot of people will be interested (or at least more interested than if the speaker starts by saying "I'm going to talk about containers for hypergraphs").
Otherwise, start by presenting a situation (or amusing anecdote) where your results would be applicable:
Suppose 11 of your spouse's charming relatives are visiting for two weeks. How do you seat everyone around the round table you had to drag out of the garage so that everyone has everyone else on each side of them at least once during the two weeks? What if you want to avoid their annoying cousin but would love to sit next to their mother-in-law more times?
Both of these were very interesting beginnings, and as the speakers continued through their talk, they kept connecting back to the original problem, explaining how their result provided easier or more convenient solutions to the basic problem. There is really no need to start with "thank you for attending". Wait with that until after, and save the title and collaborators until you've explained why everyone should care what you and some other people they don't know wrote a paper about. You have already been introduced by the organiser, so you can just dive right into the motivation.