Video wird geladen...
Video konnte nicht geladen werden
Consent✅ Runner, HIIT Deep left buttock pain Occasionally travels down posterior thigh with intermittent parasthesia & numbness lower leg & foot Brought on with pigeon stretching, yoga, as well 4 point kneeling core hip extension / adduction work 🤷♂️ Sitting main aggravating factor - isometric single leg bridging as... show more
16,387 Aufrufe • vor 8 Monaten •via X (Twitter)
8 Kommentare
Any concerns of placebo effect of image guided injections in centers of excellence?
No
V satisfying video, and injection too im sure! What was the underlying cause of IFI?
Piriformis?
What is a long term plan here? Repeated injections, surgical intervention, no further tx needed?
If you have access to cryoneurolysis, you can go centrally in QF with motor stimulation and do a 1 min cycle as for spacticity. I keep it as a backup if prolo dosen’t work. Additionally, If you do a follow up MR a couple of months down the line you’ll most often find muscle volume increase. Results are pretty surprising.
is the initial injectate being delivered to the posterior hip joint/ under capsule as well, aside from QF?
No - deliberately so. The needle isnt long enough anyway and you also cant reliably place the needle under USG into the posterior capsule
![Saturday mornings are for hip mobility routines. Free them of the tension gathered during the week. You'll be more relaxed for your weekend and will protect your joints over the long run. This routine relieves the stress gathered in your hips: Do one after the other and rest for two minutes. 1 - Cossack Squats [ 3 x 10/leg] You might need to hold onto something if you have no hip mobility. Go as far down as your body allows and feel the stretch. Watch how your range of motion improves over time by doing these exercises a few times a week. 2 - Deep Bodyweight Split Squat [ 3 x 10/leg] Keep your torso upright and drive your knee all the way forward. You will feel a stretch along the front of your hip and thigh on the back leg. Use support if you struggle with balance too much. 3 - Deep Squat Twist [ 3 x 10/side] Can you rotate so that your fingers point toward the ceiling? I've seen so many people with chronic pain get stuck way before. If you lack the mobility to twist, hold the deep squat position for thirty seconds to one minute. Grab onto something at the bottom if you can't stay in the position. 4- Seated Wide-Leg Adductor Stretch [ 3 x 30-60s] Sitting with wide legs will be enough for some people to elicit an intense stretch. // Many of the new chronic pain clients I meet should have started a routine like this one decades ago.](https://image.24vids.com/tw-1723331847927586832/ext_tw_video_thumb/1723331575629139968/pu/img/bX9LaYiliysctP9q.jpg)
![These quick workouts are for people who want to improve their postures. The exercises target the weakened foundations pulling your body out of alignment. The sessions are short yet effective. Use the workouts to raise yourself to the next level before the new year starts. Instructions: - Do the exercises as a circuit, or one at a time if you find the circuit format too challenging. - Repeat the circuit or each exercise 3 to 4 times. - The sessions are light enough that you can done daily. Follow the sequence from Day 1 to Day 7 in order. - Ideally done at the start of your day or at the end. Day 1: Kneeling T-Spine Stretch (0:06) [ 10 reps/side ] - Start: Kneeling, one hand on the floor. - Slide your straight arm under your torso. - Reach as far forward as possible. - Keep your hips squared throughout the motion. Kneeling T-Spine Rotation (0:12) [10 reps/side] - Start: Kneeling, one hand on the floor, the other behind your head. - Rotate inward to point your elbow toward the floor. - Rotate outward to point your elbow toward the ceiling. - You may not have the mobility to go all the way up. Stop when you reach your max, and do not force it. Supine Around The World (0:18) [10-20 reps/side] - Start: Lying on your back, arms by your sides, palms toward the ceiling. - Rotate your arms back to have your hands pointing behind you. Day 2 Side-Lying Quad Stretch (0:27) [30-60 seconds/leg] - Start: Lying on your side, legs stacked. - Grab your ankle and pull your foot toward you. - Use a towel or elastic band if you cannot reach your ankle. Hip Bridge, Feet on Bench (0:37) [10 reps with a 3s hold at the top of each rep] - Start: Lying on your back, feet on an elevated surface like a bench, couch, or chair. - Lift your hips off the floor to align your upper and lower bodies. - Adapt: If you cannot lift all the way, do the same exercise with your feet on the floor. - Adapt: If this variation is too easy, do one leg at a time. Dynamic Hip Extension Plank (0:49) [ 10-20 reps/leg] - Start: Push-up position. - Lift one leg as high as possible, keeping it straight. - Alternate legs. - Adapt: Rest on your forearms if you struggle to hold the push-up position. Day 3 Sky Reach (1:04) [10 reps/side] - Start: Deep Squat position - Rotate one arm to have your hand pointing toward the ceiling or as close as you can reach. - Do not force the range of motion if you block before that. - Alternate arms. - Adapt: Kneeling variation. Side-Lying External Rotation (1:21) [10 reps/arm] - Start: Lying on your side, one hand supports your head, and the other holds a dumbbell or any weighted object if you lack equipment. - Rotate the weight directly above your body while keeping your arm locked to your torso. Bent-Over Reverse Flys (1:28) [20 reps] Start: Bend forward, arms hanging down. Ensure your back is straight and your hips are held high. Lift your arms to shoulder level and keep them straight. Adapt: Add weights if bodyweight is too easy. Day 4 Deep Squat (1:36) [10 reps] - Start: Feet slightly wider than hips apart. Inhale first, then hold it in. - Lead the motion by sliding your knees forward and letting your hips naturally fold. - Go as low as your body allows, then push back up. - Exhale on the way up. - Adapt: Hold onto something stable to help you go lower. Hip Thrust (1:44) [10 reps with a 3s hold at the top of each rep] - Start: Both feet on the ground, shoulders on the edge of a bench or couch, hips down. - Lift your hips to align your lower and upper bodies, and hold. Side Bridge (1:57) [10 reps/side] - Start: Side Plank position, legs stacked. - Lift your hips off the floor and hold when you reach the top. - Ensure your head stays aligned with your shoulders and hips. Avoid bending it forward to make the motion easier. - Adapt: Bend the lower leg. Day 5 Wall Slides (2:23) [10 reps] - Start: Stand in front of a wall with your forearms and hands flat behind you. - Be as close as you can to the wall, but you can take a few steps forward if you lack the mobility to keep your arms flat. - Straighten your arms while keeping them flat against the wall. Y-Raises (2:29) [10 reps with a 3s hold at the top] - Start: Lying on your stomach, arms in a Y position, thumbs toward the ceiling. - Lift your thumbs toward the ceiling and hold. - Lower back to the floor. T-Raises (2:41) [10 reps with a 3-second hold at the top] - Start: Lying on your stomach, arms in a T position, thumbs toward the ceiling. - Lift your arms toward the ceiling and hold. - Lower back to the floor. Day 6 Seated Toe Pointing (2:56) [10 reps] - Start: Sitting on the floor with your legs ahead of you, feet together. - Point your toes forward. - Point them toward you. Seated Ankle Rotations (3:02) [10 reps] - Start: Sitting on the floor with your legs ahead of you, feet apart. - Rotate your feet like a windshield wiper, going as deep as possible on either side. Reverse Plank (3:09) [Hold for 30-60 seconds] - Start: Ball of the feet on a step, heels hanging. You may hold on to a ramp or wall for added support. - Lower your heels as low as possible and hold at the bottom. - Lift your heels as high as possible. Day 7 Unilateral Chest Stretch (3:21) [30-60 seconds/arm] - Start: Place your hand flat against the wall at shoulder level. - Rotate your body in the opposite direction and hold. - Placing your hand higher than shoulder level increases the stretch. Cat-Cow Stretch (3:31) [10 reps] - Start: Hands and knees on the floor. - Round your lower back like a cat (pulling it toward the ceiling) and hold. - Push it down as much as possible and hold. Wide Leg Adductor Stretch (3:38) [30-60 seconds] - Start: Sit on the floor with your legs as wide as possible, hands right in front of you. - Hold the position. // Start your New Year's resolutions with these quick posture workouts. You'll build up the foundations needed to withstand 2024 and crush your goals. Have an excellent weekend!](https://image.24vids.com/tw-1735744101629215154/ext_tw_video_thumb/1735743000876040193/pu/img/hZ05dLDaEqKUU5DF.jpg)
![I think I can finally report some success training a quite accurate IDM capable of recovering keystrokes from Minecraft gameplay, even in quite PvP-heavy situations. At this point the model does not only know what keys are pressed to the extent reasonably discernible, it also knows how fast it is moving in 3D space at all times, even when knockback is mixing with the self-move impulse. Now, recovering keystrokes from normal external capture footage is just about impossible. E.g. W/A/S/D does exactly nothing during partial tick frames and jumping mid-air is also equally useless, so asking the model to recover key down states is inherently unreasoanble. Mouse deltas are also completely arbitrary units, as game mouse sensitivity introduces an arbitrary scale factor into the equation. The only good option is to think carefully about your model-environment contract, and only record "logical actions", not raw keystrokes. So here's a few unfortunate lessons I had to learn in roughly this order. - Choose good units. (bad: mouse deltas, good: delta radians [yes, you will need game-internal state]) - Capture from inside the main game loop and read the game fbo to get consistent frame-action pairing. Doing post-mortem pairing is hopeless. - Carefully define when you think keystrokes actually have an effect. (jump only works on ground, when flying or in water etc.) More subtle: The key may already be down, but no tick has happened yet to actually use the value. Hence: ignore Seperate gamestate into "fast and slow-moving" components. E.g. movement is likely tick based, camera rotation is very likely updated every frame in essentially every game ever. - Think about your frame-action correspondance contract (How old is the frame in relation to the inputs you capture? Will double or tripple buffering affect you?) Think about the game loop timeline, where you are sampling, how old the data you are reading is, and where the ticks are happening around you. Language models used to simply not have a model-environment contract, but even now with the model "living" in a designated harness, the contract still boils down to formatting, and tool implementation intrinsics. While also important, it is still quite a bit more obvious because the violations are in some way shape or form reflected as text you can actually see. - ffmpeg dropping frames cummulatively screws the model the further you get into the sequence because your targets are now shifted. If you can't encode the video in real-time, too bad. - Sodium has a frames in flight system different from vanilla Minecraft, which will also offset your targets from your frames. (there goes that data...) - Models are succeptible to latency. If there is too big of a delay between action and on-screen reflection, your performance degrades. At this point I realize ~100hours of gameplay is essentially no longer usable as a dataset. You can train on this data, but all you'll get is a mushy mess. However, some good news: - Making the model predict physics gamestate scalars helps the model generalize. For instantaneous events like jump, it's unreasonable to ask the model emit a short burst of jump=true at exactly the right time, however if you also predict your current y-velocity, the model has supervision signal for the "latent" from which that onground jump becomes apparent. Recovering x/z motion is also somewhat easier than unmixing it into plausible keystrokes for inertia-heavy player controller logic. - Regressing physics gamestate scalars also seems to make your dataset "bigger". While pure keystroke classification will overfit quickly, predicting exact physics gamestate scalars forces the model to generalize more and you can tolerate far more epochs before validation loss starts to stall out. This is the only reason why it was bearable to dump 100h+ of dataset hours and replace it with ~3 hours of gameplay after the 4th revision of the file format (yeah...) and somehow still have better performance. Now, you might be asking, "isn't this brittle?" and the answer is yesn't. Frame-action correspondance matters for training, but not so much during inference. So as long as you are sampling in roughly the same interval as your training data, you aren't violating any hard contract per-se. Somewhere around the frames ticks are happening, and during training you capture various tick-capture offset relations per random chance, so nothing is too obviously wrong here. HOWEVER, you will get screwed by gui scale, shaders, resource packs, "shit that recording is 1920x1040 because somebody doesn't know fullscreen exists" and other unfortunate edge cases of reality. But I suppose this is the role of dataset size. If all those "contract violations" that a youtube video has compared to the training data are addressed, I think this is a way to turn Youtube into a labeled dataset. I could never shake the feeling that VPT is a sound idea in practice, while never having been properly executed, and I think one reason why it hasn't is because that label boostrapping part is just a pain in the butt to get right. Now, what the player is doing is of course not the only label you can extract from video, but it has to be one of the targets predicted during pretraining to "align" the pretraining objective. Some notes on the video here, the colored dots on the analog visualizer are the ground truth, while the gray dot is the model prediction. Green means correct prediction, red means incorrect prediction at that frame. Model P(key) reports how wrong the prediction is from green (0.0) to red (1.0). You will also notice that during periods of rapid slow down, left and right actions become close to irrecoverable, because there is just that little motion. And some jump actions are not predicted correctly because I got the detection condition for jump events wrong... (duh) LMB/RMB for other than sustained events (like item-consume and block break) also seem to be hopelessly irrecoverable for now. Swing was supposed to do the same thing as motion y did for jump, but its too well behaved as an increasing counter. Maybe partial-tick interpolated values work better (v5 file format then... ugh..)](https://image.24vids.com/tw-2042731804780826849/amplify_video_thumb/2042721691844120576/img/yugYQfuXXQmJpdVp.jpg)