注册并分享邀请链接,可获得视频播放与邀请奖励。

与「RGB」相关的搜索结果

RGB 贴吧
一个关键词就是一个贴吧,路径全站唯一。
创建贴吧
用户
未找到
包含 RGB 的内容
this is clinically monitored 🩺
0
12
25
0
转发到社区
兄弟们好东西来了, Qwen-Image-2.1 社区无审查 GGUF 版放出了, 安全拦截全部下线,懂的都懂。 而且作者搞了一套极度聪明的切分跑法,文本编码放 CPU,主模型进 GPU,生成几乎不降速,显存立省 9 到 17G,直接把本地出图的门槛焊死在平民卡上。 平时在各种在线平台跑图,最搞心态的就是动不动弹窗拒识或者直接给你涂黑一张图。 这个版本最爽的地方就在于把内置的安全拦截全拔了,提示词不设防,敏感内容和成人画面直接放行出图。完全是在自己电脑本地离线跑,彻底摆脱外部干预。 更绝的是它的显存优化方案。以往跑这类大底座,动不动就得吃掉 20 多 G 显存,平民显卡直接当场劝退。这次作者在 ComfyUI 里的解法极其优雅:把极其吃显存的 8B 文本编码器直接甩到电脑普通内存去跑,显卡只负责吞下 7B 的视觉生成主模型。 因为文本编码在整个出图过程中只在最开头执行一次,后续几十步的扩散迭代完全在显卡里跑, 所以这么切分不仅几乎不拖慢出图节奏,还能凭空帮显卡省出 9 到 17GB 显存。手头只有 8G 到 12G 显存的甜品卡,也能在本地顺畅跑完 2K 分辨率的大图。 除了省显存,工作流也是直接大瘦身。以前在本地搭一套能用的生图流程,先文生图,再换重绘模型修细节,最后还得外挂抠图插件去背景,工作流连线密密麻麻像盘丝洞。 Qwen-Image-2.1 直接把文生图和图像编辑焊进同一个模型里,而且原生自带 64 通道 RGBA 输出,在提示词里指定透明背景就能直接吐出带透明通道的切片素材,抠图节点全都可以删掉了。 另外它最多能吃 10 张参考图。最实用的场景就是多人合影:直接把 6 个人的单独大头照喂进去,模型能一次性合成出一张光影相互反射、透视完全协调的多人自然大合照,彻底告别以往硬拼贴的僵硬感。画面质感上的提升也很扎实,中英文字排版不再扭曲画符,皮肤质感、毛发细节和真实光影过渡非常自然,完全摆脱了早年那种油腻腻的塑料感。 既有不设限制的纯粹自由,又有极致省显存的工程优化。手头有平民卡又想折腾本地高质量出图的老哥,这波直接冲就完事了。 模型和实测避坑放在一楼。
显示更多
0
4
137
27
转发到社区
am i right?… yes or yes?
0
14
3.4K
97
转发到社区
When colors feel too real… RGB MiniLED brings every shade and detail closer to reality. Hisense. The Origin of RGB MiniLED. Hisense, Innovating A Brighter Life.
When the game reacts with you, every moment hits harder. With Razer Chroma RGB and Razer Sensa HD Haptics, supported titles can sync with your setup through reactive lighting and tactile feedback you can see and feel. Watch how it comes together:
显示更多
Codex with RGB image inputs could do this. LLM agents are better than most people thought.
Light up your voice with the Razer Seiren V3 Chroma 32-Bit DSP: Stream-reactive Razer Chroma RGB transforms your setup by enhancing your stream with immersive, dynamic lighting. Combined with a 32-Bit DSP and AI-powered Smart Audio that adapts to your space for consistently clean vocals, you can focus on entertaining your audience while your setup does the rest.
显示更多
my phone and i are merging into a single organism
0
12
18
1
转发到社区
china is not slowing down and this one is insane.. Chinese researchers just open-sourced SLAM3R, it rebuilds a full 3D scene from a single regular video, in real time. no LiDAR, no depth sensor, no camera rig. one phone-grade video in, a 3D world out. > runs at 20+ FPS on a single GPU > monocular RGB, just walk around filming > feedforward neural net, no slow pose-optimization step > beats optimization-based methods on reconstruction benchmarks > handles long walkthroughs without drifting apart what makes this work: older methods solve for camera position frame by frame, which is why they're slow. SLAM3R skips that and regresses the 3D geometry directly. that's how it hits real time. 100% free and open source:
显示更多
0
15
393
75
转发到社区
With the Razer Soma Chroma, your Razer Chroma RGB battlestation no longer ends at the desk: Immerse yourself in dynamic lighting that reacts to gameplay across over 300 supported titles, with every effect customizable through Razer Synapse on PC or the Razer Furniture app on mobile.
显示更多